Wayne State University

Wayne State University Dissertations

1-1-2017

Mechanical Interventions In Soft Tissue RepairElizabeth Marie MeierWayne State University,

Follow this and additional works at: https://digitalcommons.wayne.edu/oa_dissertations

Part of the Biomedical Engineering and Bioengineering Commons

This Open Access Dissertation is brought to you for free and open access by DigitalCommons@WayneState. It has been accepted for inclusion inWayne State University Dissertations by an authorized administrator of DigitalCommons@WayneState.

Recommended CitationMeier, Elizabeth Marie, "Mechanical Interventions In Soft Tissue Repair" (2017). Wayne State University Dissertations. 1724.https://digitalcommons.wayne.edu/oa_dissertations/1724

MECHANICAL INTERVENTIONS IN SOFT TISSUE REPAIR

by

ELIZABETH MEIER

DISSERTATION

Submitted to the Graduate School

of Wayne State University,

Detroit, Michigan

in partial fulfillment of the requirements

for the degree of

DOCTOR OF PHILOSOPHY

2017 MAJOR: BIOMEDICAL ENGINEERING

Approved By:

Advisor: Mai T. Lam, Ph.D. Date

Karin Przyklenk, Ph.D. Date

Wei-Ping Ren, M.D., Ph.D. Date

Harini Sundararaghavan, Ph.D. Date

ii

ii

ACKNOWLEDGMENTS

I would like to express my strong appreciation and gratitude to my advisor, Dr. Mai

Lam, for her continuous support and guidance throughout the course of this work. I also

would like to thank my committee; Dr. Karin Przyklenk, Dr. Wei-Ping Ren, and Dr. Harini

Sundararaghavan, for their insight and feedback on each of the tissue systems and the

nuances of their physiology and mechanics. I would also like to thank the Lam Lab; Bin

Wu and Zhengfan Xu for their instruction and guidance, as well as current members

Cameron Pinnock, Bijal Patel, Tiara Heard, and Ashley Apil. Additionally, Dr. Aamir

Siddiqui and Dr. Donna Tepper, have been instrumental in providing insight and tissue

samples crucial to this work. The Rumble Fellowship has provided significant funding for

this work and I greatly appreciate this honor and assistance. Finally, I am incredibly

grateful for the continued love, support and encouragement from my husband, Scott

Meier.

iii

TABLE OF CONTENTS

Acknowledgements ....................................................................................................... ii

List of Tables ................................................................................................................. v

List of Figures .............................................................................................................. vi

Chapter 1: Introduction ................................................................................................. 1

Skin Mechanics and Suture Technique ........................................................................ 3

Biomechanics of the Menisci and Cell Development ................................................... 7

Myocardial Infarction Injury and Wound Healing ........................................................ 16

Chapter 2: Mechanical characterization of intact and sutured human skin .......... 36

Materials and Methods ............................................................................................... 37

Results and Discussion .............................................................................................. 41

Chapter 3: Mechanical stimulation drives ASC differentiation ............................... 50

Materials and Methods ............................................................................................... 52

Results ....................................................................................................................... 58

Discussion .................................................................................................................. 69

Chapter 4: Effects of SIS patch on Inflammatory Response .................................. 76

Materials and Methods ............................................................................................... 78

Results ....................................................................................................................... 88

Discussion ................................................................................................................ 102

Chapter 5: Conclusion and Future Work ................................................................ 106

iv

References .................................................................................................................. 109

Abstract ...................................................................................................................... 127

Autobiographical Statement .................................................................................... 129

v

LIST OF TABLES

Table 1. Patient data for skin use……………………………………………………………..38

Table 2. Summary of tensile properties of skin samples……………………………………45

Table 3. Summary of media components and concentrations…………………………….53

Table 4. List of inflammatory marker genes and primer sequences……………………….82

Table 5. Means and standard deviations for neutrophil and macrophage counts………..93

Table 6. Summary of brightness difference between groups………………………………96

Table 7. Tensile measurements…………………………………………………………….101

vi

LIST OF FIGURES

Figure 1: Histological analysis of each skin layer……………………………………………..4

Figure 2: Diagram showing collagen I formation relative to stress-strain behavior……….6

Figure 3: Diagram showing physiological zones of the meniscus…………………………..8

Figure 4: Diagram of how substrate stiffness and tension activate TGF-β pathway……..13

Figure 5: Myocardial infarction and post-infarct remodeling……………………………….16

Figure 6: Overview of inflammation process post-MI……………………………………….23

Figure 7: Installation of tissue samples in Instron…………………………………………..39

Figure 8: Tensile behavior of skin under low and high strain rates………………………..43

Figure 9: Stress relaxation results……………………………………………………………50

Figure 10: Mechanical stimulation device……………………………………………………54

Figure 11: Cell morphology under different media types…………………………………..58

Figure 12: ASCs exhibit increased chondrogenic and fibrogenic gene expression……...61

Figure 13. Mechanical stimulation for 1-6 hours promotes varying expression…………..63

Figure 14. Phenotype expression varied with varying strain rates………………………...64

Figure 15. Strain rates do not affect protein level collagen expression…………………..66

Figure 16. Mechanical stimulation frequency affects phenotype…………………………..68

Figure 17. Stimulation frequency does not affect protein level collagen expression……..70

vii

Figure 18. Placement of the SIS patch on infarct area……………………………………..80

Figure 19. Diff-Quik stain of example blood smear………………………………………….84

Figure 20. PCR results plotted over time relative to infarct only……………………………89

Figure 21. Immunofluorescent images showing DAPI and CD45AB……………………...96

Figure 22. H&E and Picrosirius Red stained tissue samples………………………………97

Figure 23. Images show faint collagen presence in infarct area…………………………..99

Figure 24. Raw data graphs showing samples of data output from tensile loading…….101

1

Background and Significance

Mechanical interventions in wound healing and repair have been used since the

ancient Egyptians, who were the first documented cases of splinting broken or injured

limbs (Fess 2002). While medical practices have advanced significantly since then,

mechanical interventions have and will continue to be necessary for some injuries, at

least for the foreseeable future. With this in mind, this body of work sets to investigate

some of these mechanical interventions that are designed to promote wound healing,

repair, or even replace an injured tissue.

Traditionally, describing the effectiveness of mechanical interventions has been

simply a matter of whether or not mechanical integrity was restored to the damaged

tissue. For example, the success of the earliest splints was determined by observed

improvements in the healing of broken bones. However, mechanical interventions can

have physiological effects, both expected and unexpected, that can drastically affect other

aspects of the tissue’s performance. These physiological effects can range from providing

temporary support to damaged protein filaments, to directly affecting gene expression in

cells, to altering system-wide physiological responses to an injury. The purpose of this

work is to investigate a range of previously used mechanical interventions and determine

the physiological effects, both expected and unexpected.

Firstly, one of the most widely used and simple mechanical interventions is

investigated. The use of sutures to close wounds to promote faster healing is a practice

familiar to the general public. Human skin samples were sutured and loaded in tension in

CHAPTER 1: INTRODUCTION

2

multiple orientations to determine how the injury and suture treatment change the tensile

mechanics of our most superficial organ. Next, the cellular component of tissue

engineering was addressed. Specifically, a unique biochemical and mechanical approach

was take to create a meniscus cell source from human adipose-derived stem cells.

Thirdly, the inflammatory (biochemical) effects of decellularized ECM in a wound healing

model were analyzed. A commercially available decellularized material, porcine small

intestine submucosa (SIS), was evaluated as a patch treatment in a rat myocardial

infarction model. While the mechanical benefits of an SIS patch treatment have been

established, it was previously unknown how this decellularized tissue would affect

inflammation, and therefore wound healing, of such a critical injury. While each of these

individual works focuses on a different tissue system, each tissue system presents with

specific injuries that can be aided by mechanical interventions. These tissue systems and

mechanical interventions are described, tested, and discussed in the following pages.

Skin Mechanics

The integument system, also known as skin, is a three-layered organ that covers

the entire body and serves as a physical barrier to protect the body from its surrounding

environment. To function properly, skin must be strong and durable but flexible enough

to allow movement. This is accomplished through an extracellular matrix (ECM),

comprised of vast networks of collagens and elastin fibers as well as a complex population

of glycosaminoglycans (GAGs) and proteoglycans (Corr 2011). While all components are

present in each of the three layers, it is the largest layer, the dermis, that bears the

greatest responsibility for the skin’s mechanical behavior. Several types of collagen are

present in the skin, although collagen I has the greatest direct contribution to tissue

3

mechanics, and therefore will be the main focus of this analysis. Collagen I in its mature

form consists of globular collagen units covalently bonded into a right-hand triple helix

filament (Holzapfel et al. 2000). Collagen I is very strong and stiff when stretched along

its long axis, with an elastic modulus in the range of 5- 10 GPa (Wenger et al. 2007). By

dry weight, collagen I accounts for 60-80% of skin, and therefore is the dominating protein

in skin’s mechanical behavior (Holzapfel et al. 2000). Elastin is another fibrous ECM

protein and it accounts for 5-10% of skin by dry weight (Holzapfel et al. 2000). Elastin

differs from collagen I in that its fibril structure is seemingly more organized. This

organization is key to its mechanical behavior. Under tension, the fibrils display entropic

straightening, resulting in energy “absorption” in the filament (Holzapfel et al. 2000). The

elastin fibers have various dominating orientations depending on the location of the skin

on the body, yet is always aligned with the layers of the skin (Holzapfel et al. 2000). Also

present in the skin are proteoglycans, the most common of which is versican (Carrino et

al. 2011). Proteoglycans are heavily glycosylated proteins that act to retain water in the

tissue (Carrino et al. 2011, Holzapfel et al. 2000). In tensile skin mechanics,

proteoglycans act as lubricants easing movement of collagen filaments past each other

as tension is applied (Carrino et al. 2011, Holzapfel et al. 2000). This unique combination

of ECM proteins gives rise to viscoelastic mechanical behavior, in that the tissue behaves

in a time-dependent fashion when under load (Corr et al. 2011). Under non-loading

conditions, collagen I fibrils are crimped and disorganized among proteoglycans and

elastin fibrils (Holzapfel et al. 2011). The crimped, random organization of collagen can

be seen in Figure 1, along with the organization of elastin fibrils.

4

When a load is applied in tension, the ECM proteins undergo three distinct phases.

The first phase involves primarily the uncoiling and straightening of elastin filaments. The

absorption of energy by elastin filaments results in minimal strain, or lengthening, of

collagen and of the tissue as a whole. Rather, the elastin filaments straighten and stretch,

preserving the overall integrity of the tissue (Holzapfel et al. 2000, Langdon et la. 2017).

This phase I behavior is frequently referred to as the “toe” region on a stress-strain curve.

Beyond this initial phase, the mechanical behavior of skin is primarily dominated by the

behavior of collagen (Corr et al. 2011, Holzapfel et al. 2000). This is illustrated in the

Figure 1. Histological analysis of each skin layer. Areas of cellularity and structural

variation can be seen each each of the three layers (A-C). High collagen content and its

disorganized structure can be seen in D-F. Elastin’s orientation can be seen in the cross-

sectional view, particularly in the epidermis (G) and dermis (H). Scale bar = 200 μm.

5

second phase, in which elastic behavior is observed and the stress-strain relationship is

mostly linear. In this phase, the crimped collagen I fibers are stretched and straightened,

aided in part by lubrication from proteoglycans (Holzapfel et al. 2000, Corr et al. 2011).

As the collagen fibers are merely reoriented and not deformed, they will resume their

original, crimped form if the stress is removed, exhibiting elastic behavior. Finally, in the

third and final phase, collagen fibers are stretched, deformed and eventually break and

slide past each other (Holzapfel et al. 2000). This is the plastic deformation region and it

concludes when the collagen fibers break, signifying failure strength has been reached.

All of this behavior is captured in Figure 2.

While this behavior is very consistent in uninjured skin samples, it is largely

unknown how mechanical interventions, such as suturing may affect this behavior. The

vast majority of skin mechanics studies involve computer models, xenografts, or both

(Chanda et al. 2016, Levi et al. 2016, Capek et al. 2012). The few studies that have

investigated suture mechanics in human skin have done so in vivo, which clearly limits

the mechanical parameters that can be applied. As patient consent and comfort much be

considered in in vivo work, these studies are limited to low tensile loads and strains, it is

unlikely that they encompass the full mechanical behavior of sutured skin that a patient

could experience. For example, one such study limits the tensile strain to 5%, which is

well below the failure strain of intact human skin, which has been reported to be as high

as 115% (Levi et al. 2016, Holzapfel et al. 2000). It is therefore highly likely that the

reported mechanical behavior in such studies are restricted to the viscoelastic toe region

of skin’s stress-strain curve (Figure 2). However, skin is not limited to 5% strain in day-to-

6

day activities and therefore this information cannot be considered sufficient to

characterize the full range of tensile mechanics of skin.

Sutures are perhaps the most commonly used and most basic mechanical

intervention in regenerative medicine. While numerous materials are used for sutures, as

a general rule, they have relatively high stiffness and strength, in the range of 5-50 GPa

(Greenwald et al. 1994). It is unclear how sutures affect skin mechanics, as there is little

to no research on the topic. As the skin’s function is dependent upon its integrity, it is

highly important that we understand how suturing affects the integrity, and therefore

function, of skin. While numerous attempts at creating synthetic skin and computer

models have been attempted, accurate information is still lacking (Chanda et al. 2016,

Capek et al. 2012). Therefore, part of

this body of work will focus on how

sutures affect skin mechanics.

Biomechanics of the

Meniscus and Cell Development

Like the skin, the meniscus is

a complex tissue designed to

withstand a range of forces that

occur in everyday movements.

Unlike skin however, the meniscus

does not self-heal, even with the

assistance of basic interventions,

such as sutures. Meniscal injuries are among the most common knee injuries

Figure 2. Diagram showing collagen I fiber

conformation relative to stress-strain behavior.

Phase I (“toe region”) minimally affects collagen

fibers. Phase II shows the transition from viscous to

elastic behavior. Phase III shows the purely elastic

region leading to the plastic region (Holzapfel 2000).

7

experienced by children and adults in the United States (McDermott et al. 2008, Malcris

et al. 2004, Buma et al. 2004, Baker et al. 2009, Thambyah et al. 2006, Pteri et al. 2012).

The meniscus is a disc-shaped tissue that lies on the surface of the tibial plateau in direct

contact with the femoral condyles (McDermott et al. 2008, Malcris et al. 2004, Buma et

al. 2004, Baker et al. 2009, Thambyah et al. 2006). The meniscus is a connective tissue

that helps bear and distribute loads experienced by the knee during physical activity,

preserving the tibial articular cartilage underneath. Injuries to the meniscus generally take

the form of an abrasion or a tear and most commonly occurs during a rapid, dynamic

movement in which torsion is applied to the meniscal surface (McDermott et al. 2008).

Limited vasculature and exposure to high loads prevent native tissue from self-healing

and current treatments are mainly restricted to partial or complete menisectomies

(meniscus removal), which put the patient at very high risk for degeneration of articular

cartilage and osteoarthritis (McDermott et al. 2008, Malcris et al. 2004, Buma et al. 2004,

Baker et al. 2009, Thambyah et al. 2006, Son et al. 2013). Damage of the meniscus leads

to increased compressive and tensile stresses on the articular cartilage and often leads

to premature osteoarthritis and possible joint replacement (Buma et al. 2004). For this

reason, a treatment beyond menisectomy is required. Biomaterial and tissue engineering

strategies have been proposed for both repair and replacement of damaged meniscus

tissue, but have so far failed to address the intricacies of this complex tissue. Biomaterial

strategies have looked at allografts, natural scaffolds, and polymer-based synthetic

replacements, all of which have their advantages and disadvantages. The meniscus has

also been studied for possible tissue engineering solutions, with a major focus on

identifying a cell source to produce adequate extracellular matrix to perform the

8

mechanical tasks required of this tissue. This review will address the characterization of

the physiology and function of the native tissue and describe some of the current

techniques and research aimed at addressing additional treatments beyond

menisectomy.

Anatomy and Physiology

The meniscus is a highly specific tissue, responsible for bearing and redistributing

loads across the knee (McDermott et al. 2008, Makris et al. 2004, Son et al. 2013). On

multiple levels, the meniscus can be divided into regions based on physiology and

function. On a macroscopic level, the meniscus can be divided into three regions based

on vasculature, with the greatest vasculature, “red zone”, developing in the most lateral

third of the meniscus (Makris et al. 2004, Buma et al. 2004, Son et al. 2013). Moving

medially, the next region is a semi-vascular region commonly referred to as the red-white

zone. The inner third is almost completely avascular and is known as the white zone. All

of this is visually described in Figure 3. Damage to the outer, vascular, red zone tissue

Figure 3. Diagram showing zones of the meniscus (McDermott et al. 2008)

9

can self-heal, but damage to the inner red-white or white avascular areas is generally

stagnant due to the limited blood and nutrient supply (Makris et al. 2004, Buma et al.

2004). This limited vasculature is something of a double-edged sword as it indicates that

establishing vasculature is not a major obstacle for tissue engineering strategies, as is

almost always the case with other tissues, however the body’s capacity to heal is the

cornerstone of tissue engineering and regenerative medicine.

Beyond the vasculature, the meniscus structure is highly complex, consisting of

two distinct regions of collagen fiber arrangement. The outer third shows circumferentially

oriented collagen fibers, primarily collagen I, and experiences mostly tensile deformation

under loading. These fibers help conserve the meniscus shape and contribute to resisting

deformation due to the high mechanical loads (McDermott et al. 2008, Makris et al. 2004).

Conversely, the inner two thirds of the meniscus display radially oriented collagen I and

II fibers and experiences mainly compressive deformation under loading. These collagen

fibers are accompanied by a high concentration of proteoglycans that contribute to the

region’s compressive stiffness through water retention (McDermott et al. 2008, Makris et

al. 2004, Nishimuta et al. 2012). This complex 3D extracellular matrix structure is

designed to withstand high cyclic loads. The tissue itself is composed of a gradient of

cells that exhibit a mixed phenotype, with characteristics similar to cartilage and ligament.

These cells are commonly called fibrochondrocytes (McDermott et al. 2008, Makris et al.

2004, Buma et al. 2004, Son et al. 2013, Kun et al. 2012, Nishimuta et al. 2012). As would

be expected, meniscal cells secrete a variety of extracellular matrix proteins, including

collagens I and II, versican, aggrecan, and other glycosaminoglycans and proteoglycans.

Cells in the outer region have a fibroblast-like morphology and show increased expression

10

of collagen I and versican (McDermott et al. 2008, Makris et al. 2004, Buma et al. 2004).

Cells in the inner region have a rounded morphology and behave more like chondrocytes

including aggrecan and collagen II (McDermott et al. 2008, Makris et al. 2004). The

cellular expression and distribution contribute to a meniscus that is 72% water and 28%

organic matrix (Makris et al. 2004). Of this 28%, collagen makes up the majority (75%),

then GAGs (17%), the remaining consists of cells, adhesion glycoproteins, and elastin

(Makris et al. 2004). The location and population of meniscal cells, proteins, and collagens

in the meniscus are critical to its mechanical function. Also of note are the anchoring

ligaments that help hold the meniscus in place. The anterior intermeniscal ligament

connects the anterior horns of the medial and lateral meniscus, forcing these two

seemingly separate tissues to act as one (McDermott et al. 2008, Makris et al. 2004). All

of these physiological features contribute directly to the mechanical integrity of the tissue

and ultimately, the knee itself.

Stem Cells in Meniscus Tissue Engineering

When native tissues are incapable of providing adequate cell populations for

treatments, researchers look to other donor sources and autologous stem cell sources.

When dealing with connective tissues, such as the meniscus, there are several

autologous stem cell sources. While human embryonic stem cells (hESCs) remain the

gold standard for any stem cell-based therapy, meniscus tissue engineering strategies

have often used mesenchymal stem cells (MSCs), generally from bone marrow, although

adipose-derived stem cells (ASCs) have been gaining traction due to ease of harvest and

patient compliance (Makris et al. 2004, Baptista et al. 2013, Kokai et al. 2014, Pak et al.

2014). MSCs have been shown to readily differentiate to various connective tissues,

11

including fibroblasts and chondrocytes, although a definitive differentiation protocol to the

hybrid fibrochondrocyte phenotype has yet to be established. Similarly, ASCs have been

shown to be capable of differentiating to cartilage and fibroblasts, with evidence

suggesting that meniscus cells, or meniscus-like cells, also being very possible (Kokai et

al. 2014). Many of these studies have provided much evidence to support that use of

certain growth factors used in chondrogenesis, such as TGF-β, can induce expression of

GAGs and biglycans in meniscal cells and MSCs in culture (Buma et al. 2004). Less well-

known, platelet-derived growth factor (PDGF) has shown to counteract production of α-

smooth muscle actin, which is often stimulated by TGF-β (Buma et al. 2004). Smooth

muscle actin is a key indicator of scar tissue formation and can cause contraction of the

scaffold, often as great as 50%, which is unacceptable for a tissue construct dependent

upon its structure to fulfill its function (Buma et al. 2004). One of the more interesting

findings is that culturing ASCs in chondrogenic media can yield cells with the same gene

expression of Sox9, aggrecan, and collagen II as cartilage progenitor cells in the same

media (Baptista et al. 2013). This indicates an unprecedented affinity for ASC

differentiation to a cartilage-like phenotype and gives a strong basis for cartilage and

meniscus tissue engineering techniques involving ASCs.

Beyond biochemical factors, mechanical stimulation of stem cells can encourage

differentiation, especially to cell types that typically undergo high and/or cyclic loading

conditions, such as muscle, ligament, bone, and cardiac tissue (Park et al. 2013). While

the exact pathways are unknown, it is understood that transduction of mechanical signals

are responsible for the up regulation of these proteins in the cell (Schwartz et al. 2013).

During development, these mechanical stimuli can be the result of external forces on the

12

embryo, or forces generated from differences in tissue growth rates and/or early

contraction of developing muscle and neural tissue (Schwartz et al. 2013). In fact,

restricting movement in developing chick embryos results in malformation of bone and

limited formation of cartilage and other connective tissues (Schwartz et al. 2013, Mikic et

al. 2000). It is also known that mechanical forces are required for proper development of

non-mechanical tissues such as the vasculature, kidneys, and lungs, supporting the need

to investigate stem cell differentiation techniques beyond biochemical factors (Schwartz

et al. 2013).

One of the more thoroughly studied effects of tensile strain is its effects on the

TGF-β pathway (Meier et al 2016, Li et al 2010). An illustration of the mechanism can be

viewed in Figure 4. Activation of this pathway results in activation of numerous key

transcription factors within the cell (stem cell or adult cell), leading to changes in gene

transcription and, often, phenotype in the case of stem cells (Li et al 2010). Activation of

TGF-β pathway is crucial for differentiation of fibroblasts, chondrocytes, and osteocytes

(Li et al 2010, Khani et al 2014, Saha et al 2008). Specifically, activation of TGF-β pathway

leads to up regulation of collagen I and versican (Li et al 2010, Khani et al 2014). Uniaxial

tension has also been shown to align F- actin and other skeletal components, increasing

the elastic modulus of the cell and altering its transcription pathways (Khani et al 2014,

Teramura et al 2012). Importantly, tensile strain applied to stem cells has been shown to

inhibit certain differentiation pathways, particularly adipogenesis (Sen et al 2011). This is

achieved through changes in the cytoskeleton which not only increase stiffness of the

cell, but increase focal adhesions, which also affect the cell’s transcription pathways

(Teramura et al 20102, Sen et al 201). Additional focal adhesions make the cell

13

increasingly sensitive to tensile strain, effectively initiating a feed-forward mechanism

towards differentiation to a musculoskeletal phenotype, such as fibroblasts, osteocytes,

or skeletal muscle (Sen et al 2011). The specific differentiation pathway is dependent

upon other factors in the cells’ environment at the time of applied tension.

Numerous studies have shown that substrate stiffness can direct stem cell

morphology, with softer substrates encouraging a more rounded morphology in MSCs

and harder substrates directing cells towards elongation (Galie et al. 2013, Cheung et al.

2009, Lam et al. 2009). Healthy chondrocytes in the knee in vivo are generally subjected

to repetitive cyclic loads of 3-20MPa, which can encourage increased expression of

Figure 4. Diagram of how substrate stiffness and applied tension result in activation

of TGF-β pathway. Expressed proteins act to resist applied tension (Wells et al. 2008).

14

GAGs and collagen II (Buma et al. 2004). Similar stimulation can affect differentiation of

MSCs, especially in the presence of chondrogenic media including TGF-β (Petri et al.

2012, Kun et al. 2012, Nishimuta et al. 2012, Baptista et al. 2013, Buma et al. 2004).

Fibroblasts can similarly be obtained, generally through tensile strains of 10-20% (Buma

et al. 2004). It stands to reason then, that combinations of these growth factors and

mechanical stimulation techniques could encourage features of both fibroblasts and

chondrocytes in MSCs. It should be noted that the term mechanical stimulation can apply

to a wide range of techniques and magnitudes of stimulation. One such study looked at

combining continuous perfusion with 8 hours of cyclic compression. MSCs seeded on a

collagen I fibrous scaffold were exposed to perfusion with chondrogenic media, perfusion

with compression, or no stimulation at all. Compressive modulus of scaffolds that

underwent perfusion and mechanical stimulation doubled that of the control group (24.7

kPa and 12.3 kPa, respectively). A significant increase in the rate of production of matrix

proteins was also observed, yet levels were well below that of native tissue. It is also

important to note that even the highest compression modulus is still a far cry from the

native modulus, about 400 kPa, partially due to the relatively low compressive modulus

of collagen I and limited time allowed for cells to replicate and produce matrix (10 days)

(Petri et al. 2012, Lam et al. 2009). Another example of mechanical stimulation used

MSCs with a fibrin and alginate hydrogel. By varying the concentrations of each

component, researchers were able to control the differentiation and scaffold

characteristics. Fibrin helps maintain gel extensibility and encouraged cell proliferation.

Alginate encourages expression of chondrogenic genes, such as Sox 9 and Aggrecan,

as well as production of collagen II and GAGs (Kun et al. 2012). It also resists the

15

contraction and shrinking of fibrin, although it is generally brittle and can tear easily (Kun

et al. 2012). Based on previous work demonstrating that tensile strain increases GAG

and collagen population, 1 week was found to be appropriate duration for tensile loading

for optimal expression of key meniscus genes in the fibrin-alginate gel. Through

measuring mechanical and biochemical properties and gene expression, they determined

that a 40:8 ratio of fibrin to alginate had the best combination of net effects and would be

the best composition of the two for future applications (Kun et al. 2012). This study is one

of the few that combines that benefits of stem cells and adjustable materials with

mechanical stimulation, an approach which will most likely be required in a long-term

clinical solution. A third study looked at the effect of strain rates on healthy cartilage and

meniscus explants to compare published effects of strain on developing cells and reaction

of healthy cells. Cartilage and meniscus explants were subjected to several compressive

strain rates (0.5%/s, 5%/s, and 50%/s) and tested for metabolic activity, water, and GAGs.

Total strain was 40% for all strain rates. At the higher loading rates, cells began to lysis

and no significant GAG increase was observed for any of the strain rates (Nishimuta et

al. 2012). This is to be expected as normal strain rates for healthy adults is generally 19-

21% (Chia et al. 2008). Loading beyond 25% strain can cause permanent deformation

due to the tissue’s inability to achieve full hysteresis (Chia et al 2008). This indicates that

relatively higher strains than in vivo may be necessary for full differentiation in vitro.

Myocardial Infarction Injury and Wound Healing

When considering the role of mechanics in physiology, it is impossible to overlook

the cardiovascular system, in which the heart serves as a pump to circulate blood

throughout the body. One of the more detrimental injuries to this system is myocardial

16

infarction (MI), which can lead to heart failure and death (Zhang et al. 2015). Like the

meniscus, the heart has limited capacity to heal. Unlike meniscus, but similar to skin,

relatively simple mechanical interventions can vastly improve long term outcomes. MI

occurs when an occlusion is formed in a coronary artery, restricting blood flow to a region

of cardiac tissue. An illustration of this can be seen in Figure 5A. The induced ischemia

results in localized cell and tissue death, with a resulting scar tissue forming over the

injured area (Olivetti et al. 1990). This layer of scar tissue is stiff and exerts additional

forces on the surrounding, surviving, myocytes as the ventricle contracts and distends.

This cyclic tensile strain results in a condition sometimes referred to as tissue “slippage”,

in which the surrounding tissue is stretched and the scar tissue is thinned. Initial slippage

can occur as early as a few days after the initial infarct (Olivetti et al. 1990). This tissue

remodeling increases the total volume of the ventricle while decreasing contractility of

surviving myocytes (Olivetti et al. 1990) In rare cases, this can lead to scar rupture, but

more frequently leads to long-term remodeling and complications, such as heart failure.

A visual representation of this process can be seen in Figure 5B.

Figure 5. Myocardial infarction and post-infarct remodeling. (A) Myocardial infarction (MI)

occurs when a coronary artery is obstructed (1). The ischemia results in tissue death (2) (Wan et

al. 2013). (B) Images showing infarcted rat heart (left), two patch treatments (middle), compared

to healthy tissue (right) (Wan et al. 2016).

A

B

17

The formation of this scar tissue and its mechanical properties are the key to

determining long term prognosis (Wan et al. 2013, Zhang et al. 2015). Tissue remodeling

of the infarct area is dependent upon local inflammatory response, with “normal” healing

resulting in a thin layer of scar tissue that cannot properly resist the wall pressure and will

eventually distend (Zhang et al. 2015). If the scar does not provide resistance to these

high pressures, it will increase wall stress on the surrounding tissue. Increased wall stress

can affect cell metabolic processes, which leads to increases in the dysfunction and

remodeling in the area (Zhang et al. 2015). The resulting complex tissue remodeling is

dependent upon local and systemic inflammation and excessive inflammatory response

can result in poor remodeling that increases the risk of development of heart failure (Wan

et al. 2015). However, impeding the inflammation response can result in non-adaptive

fibrosis, which can lead to cardiac rupture (Wan et al. 2015). Clearly this indicates that

the components of the inflammatory response can directly predict patient outcome. A high

level overview of the inflammation response can be seen in Figure 6, along with the follow

written description.

Inflammation Response in MI

Upon the onset of ischemic conditions, hypoxic cardiomyocytes release danger-

associated molecular pattern (DAMP) molecules (Altara 2016, Boag 2016, Feng et al.

2015). This class of molecules include DNA, heat shock proteins, and adenosine

triphosphate, among others (Altara 2016, Feng et al. 2015). The release of these

molecules trigger the complement system cascade and signal neutrophils to the injured

area (Altara 2016, Boag 2016, Feng et al. 2015). DAMPs can activate toll-like receptors

(TLRs) in other cells, leading to cytokine release (Boag 2016, Feng et al. 2015). Damage

18

to myocytes also results in release of reactive oxygen species (ROS), which can cause

further direct industry as well as stimulate cytokine release (including TNF-α and IL-1β)

(Frangogiannis 2002, Feng et al. 2015).

Injury to myocytes also activates the complement cascade via release of

mitochondria membrane components that activate C1-C4 of the cascade (Frangogiannis

2002, Feng et al. 2015). Activation of the complement cascade contributes to leukocyte

recruitment (Frangogiannis 2002, Feng et al. 2015). C5a has been shown to be needed

for local monocyte chemotaxis to the injury area (Frangogiannis 2002, Feng et al. 2015).

TNF-α is a known inducer of the cytokine cascade (Frangogiannis 2002). ROS and C5a

are thought to induce Mast cell degranulation (Frangogiannis 2002). ROS can also

promote neutrophil binding to endothelial cells and promote chemotaxis (Frangogiannis

2002).

Ischemic cells also release ROS and trigger release of cytokines, including TNF-

α, IL-6 (Boag 2016, Frangogiannis 2016). Cytokines released, including TNF-α and IL-

1β, activate the nuclear factor κβ (NF-κβ) complex (Frangogiannis 2002). TNF-α is

thought to have other roles in inflammation post-MI (Frangogiannis 2002). TNF-α receptor

knockout mice had significantly larger infarcts, greater myocyte necrosis, and reduced

outcome compared to control animals (Frangogiannis 2002). This suggests that TNF-α

may be necessary to promote and complete inflammation, likely due to activation of NF-

κβ, which is known to be active in inflammation and inflammation-related pathologies

(Lawrence 2009).

Neutrophils in Inflammation

19

Along with increased matrix production, these initial inflammatory pathways have

an important role in recruitment of leukocytes to the area. IL-1β importantly acts to

stimulate neutrophil migration into the area (Chen et al. 2013). As the first leukocyte to

“arrive” to an injured area, neutrophils are an excellent indicator of the time progression

of inflammation. Indeed, the beginning of inflammation is marked with a sudden influx of

neutrophils to the area, especially after reperfusion (Stuart et al. 2016). Activation of the

complement cascade, ROS, and TNF-α trigger circulating neutrophils to begin to arrive

at the injury site (Bonaventura 2016, Boag 2016). Upon arrival, chemokines released from

the injury help neutrophils adhere to the endothelial cells in the surrounding blood vessels,

which contract and reduce the number of interendothelial junctions to permit the

neutrophil passage (Bonaventura 2016, Boag 2016). Upon passage, they release

chemokines, cytokines, inflammatory mediators, and reactive oxidative species (ROS).

Cytokines and chemokines released initially include TNF-α, monocyte chemoattractant

protein-1 (MCP-1), IL-6, IL-2 (Altara 2016, Bonaventura 2016, Stuart 2016). In particular,

TNF-α expression and concentration directly correlates with patient mortality (Feng et al.

2015). Within a few hours, neutrophils are present in the injured area and last for the first

few days of inflammation (Caimi et al. 2015, Stuart et al. 2016). In the injury area,

neutrophils act to phagocytose the injured cell debris (Stuart et al. 2016).

These initial actions lay the groundwork for the overall inflammation process,

particularly influencing ECM turnover and macrophage response and behavior (Altara

2016). After adhesion, neutrophil granules release MMPs, particularly MMP-2, MMP-8,

and MMP-9, enabling neutrophil migration further into the tissue (Altara 2016). Matrix

metalloproteinases (MMPs) are a family of zinc-dependent peptidases that actively cleave

20

various ECM proteins (Iyer 2016). MMPs are released by a number of cells in the heart

including cardiomyocytes, macrophages, fibroblasts, neutrophils, and endothelial cells

(Iyer 2016). Of particular importance are collagenases MMP-1 and -8 as well as MMP-2,

which is a gelatinase (Iyer 2016). Once located in the injured area, the neutrophils release

azurophilic granules, which are visible in differential-quick stains, and include serine

proteases and neutrophil elastase (Altara 2016).

Neutrophil population dramatically decreases by day 7 as a result of apoptosis and

phagocytosis. In chronic inflammation, neutrophil population is constantly replenished as

chemokines continue to trigger attraction and migration to the injury area (Altara 2016).

Indeed, neutrophil population post-MI has been identified as a clear predictor of patient

death (Feng et al. 2015). Due to this, neutrophil population is frequently used as a tool to

identify chronic inflammation resulting from high-risk myocardial infarction injury (Altara

2016, Stuart 2016, Feng et al. 2015).

Macrophages in Inflammation

Within the first hours of injury, neutrophils and monocytes are active at the injury

location (Stuart 2016). The monocyte population increases are especially noticeable in

ischemia-reperfusion injuries, in which perfusion occurs in fewer than 90 minutes. In these

cases, circulating cells have direct access to the injury (Boag 2016). Macrophage

response occurs in two phases: local monocytes in the tissue and circulating monocytes

that arrive and differentiate into macrophages around day 3 of inflammation (Altara 2016).

Both localized and circulating monocytes arriving in the injury area differentiate to

macrophages. Macrophages exist in two distinct phenotypes (M1 and M2) (Wan et al.

2015). M1 macrophages are pro-inflammatory, while M2 macrophages are anti-

21

inflammatory and may have protective benefits post-remodeling (Wan et al. 2015).

Polarization towards one phenotype or the other is determined, at least partially, by the

inflammatory state of the injury. M1 differentiation is stimulated by high levels of IFN-γ, a

result of the complement pathway, while IL-4 is required for the M2 phenotype (Geissman

et al. 2010, Wan et al. 2015). Monocytes presenting shortly after ischemia differentiate to

M1 phenotype (pro-inflammatory) macrophages, due to the active complement system

and active cytokines. By the third day after injury, the M1 macrophages are the dominant

cell type in the inflammation area (Stuart 2016, Feng et al. 2015). M1 phenotype increases

the inflammatory process by releasing cytokines, including TNF-α, IL-1β, and MMPs

(Altara 2016, Stuart 2016, Wan 2016). M1 macrophages also produce collagenases

including matrix metalloproteinases (MMPs) (Stuart et al. 2016, Feng et al. 2015). This

activity leads to fibroblast production, migration, and ECM degradation (Altara 2016).

Additionally, M1 macrophages phagocytose myocytes and neutrophils, an action that is

thought to trigger M2 (Altara 2016).

Later in the inflammatory phase (approximately day 5 and on) M2 macrophages

are dominant, helping to induce the myofibroblast phenotype in cardiac fibroblasts, in part

due to release of TGF-β (Stuart 2016, Feng et al. 2015). Stimulating myofibroblasts

results in production of extracellular matrix, notably collagen, and scar formation (Stuart

2016). M2 macrophages increase expression of TGF-β1, encouraging the myofibroblast

phenotype in CFs and increase matrix production (Geissman et al. 2010, Wan et al. 2015,

Stuart et al. 2016). M2 macrophages have an anti-inflammatory or inflammation resolution

phenotype, expressing anti-inflammatory markers that encourage collagen synthesis,

myofibroblast phenotype, and angiogenesis (Altara 2016). Inflammation is resolved with

22

the activity of the M2 macrophage phenotype, which upregulates TGF-β, IL-10, and pro-

fibrotic lipid mediators (Altara 2016). Apoptosis of macrophages indicates the end of the

inflammation phase and beginning of the proliferation phase (Stuart et al. 2016).

Monocyte differentiation to macrophages and dendritic cells can also shed some

light on the long term prognosis in MI. Monocyte-derived macrophages are the

predominant inflammatory cell after approximately three days after injury occurs (Stuart

et al. 2016). It is theorized that ideal wound healing will include a large portion of anti-

inflammatory M2 macrophages compared to the pro-inflammatory M1 macrophages

(Wan et al. 2015). Reduction in M2 macrophage recruitment, along with lower CD4+ T

cells has been associated with weak, nonfunctional scar formation (Geissman et al. 2010,

Wan et al. 2015). Furthermore, monocytes also possess the capability to differentiate into

fibroblast-like progenitors in environments rich in TGF-β and other factors present with

the M2 macrophage (Feng et al. 2015). Therefore, the domination of the M2 phenotype

may also help increase the scar tissue formation by directing monocyte differentiation

towards a fibroblast-like phenotype (progenitor) (Feng et al. 2015).

23

Figure 6. Overview of inflammation process post-MI.

24

Cardiac Fibroblasts and Adoption of the Myofibroblast Phenotype

Like all injuries, cardiac ischemia due to myocardial infarction triggers an

inflammatory response that helps process the damaged tissue and promote wound

healing. Cardiac fibroblasts (CFs) play a significant role in directing this inflammatory

response. Unlike their myocyte counterparts, CFs are relatively unharmed by temporary

hypoxic conditions, and therefore generally survive the ischemic incident (Chen et al.

2013, Shinde et al. 2014). CFs are implicated in directly activating the molecular pathways

that stimulate inflammation (inflammasome) (Chen et al. 2013).

As injured cardiomyocytes release DAMPs, they activate toll-like receptors (Boza

2016, Altara 2016). Toll-like receptors (TLRs) are a type of IL-R ligand that can bind

DAMPs, resulting in changes to cellular behavior (Boza 2016, Altara 2016). TLRs are

present on cardiac fibroblasts, and upon binding, ERK1/2 and PI3k-Akt kinases activate

the NF-κβ pathway, resulting in transcription of pro-inflammatory cytokines as well as α-

smooth muscle actin (α-SMA) and other genes indicative of the myofibroblast phenotype

(Boza 2016). This activation also triggers activation of the complement pathway and

production of reactive oxidative species (ROS) (Chen et al. 2013). These products have

a very strong effect on cardiac cells in the infarct area, triggering a complex inflammatory

cascade (Chen et al. 2013, Shinde et al. 2014). Both the circulating cytokines and reactive

oxidative species (ROS) can act upon CFs stimulating p38k and ERK1/2 pathways (Chen

et al 2013, Shinde et al 2014, Boza et al 2016). These activated pathways, along with

mechanical stress and exposed fibronectin, act to alter the cells towards an anti-apoptotic,

myofibrotic phenotype, with increased fibrosis and expression of α-SMA (Chen et al.

2013). This new phenotype is only observed in the injured, inflammatory state. The

25

myofibrotic phenotype is primarily responsible for increased release of

metalloproteinases (MMPs) and inhibit production of MMP inhibitors (Chen et al. 2013).

This phenotype is also affected by latent TGF-β from earlier pathways, increasing the

expression and production of matrix proteins, particularly collagen (Chen et al. 2013).

This progression towards the myofibroblast phenotype simultaneously influences,

and is influenced by, neutrophil and macrophage activity. During early inflammation,

cardiac fibroblasts proliferate and migrate into the injured area, where they release MMPs

and inflammatory cytokines to contribute to the inflammatory process, along with

neutrophils (Altara 2016). TGF-β also upregulates TL4, a type of TLR, further encouraging

the myofibroblast phenotype (Boza 2016). Expression of ROS and cytokines contributes

to neutrophil infiltration in the area, while myofibroblast activity influences macrophage

phenotype (Boza 2016, Altara 2016). Approximately three days after injury, cardiac

fibroblasts in the injury area have fully adopted the myofibroblast phenotype, in parallel

with arrival of M2 macrophages (Feng et al. 2015). This phenotype preferentially

expresses TGF-β, α-smooth muscle actin, and collagen I and is responsible for scar

formation and contraction during inflammation and maturation (Altara 2016).

Mechanical stress in the area also contributes to the differentiation to the

myofibroblast phenotype. The increased tension on the cells changes the cytoskeletal

structure of the myofibroblasts (Boza 2016). This tension leads to increases in focal

adhesions, which along with the changes in cytoskeletal structure, leads to increased

expression of collagen (Khani et al 2014, Teramura et al 2012, Boza et al 2016). Collagen

expression continues until the stress experienced by the cell is balanced by the newly

formed collagen scar (Boza et al 2016, Altara et al. 2016). However, if the inflammatory

26

process does not progress properly or is not adequately resolved, complications can

occur.

Complications from Improper Inflammation

The inflammatory process in many animals, including humans, has evolved to

place priority on speed of wound closure, rather than regeneration (Godwin 2016). While

this is ideal for limiting infection and other complications, it does prevent complete tissue

regeneration after myocardial infarction. How this inflammatory process unfolds can

directly affect patient outcome post- MI (Boag 2016, Chen et al 2013).

The outcome of inflammation is twofold: clear the injured area of cell and tissue

debris, and produce the biochemical progenitors of fibrosis and angiogenesis

(Bonaventura 2016). In many injuries, including MI, resolution of inflammation occurs

seamlessly. However, unresolved inflammation can result in inadequate or excessive

scar tissue deposition, arrhythmias, and other complications in MI (Altara 2016).

No effective treatment for myocardial ischemia has been identified to date (Boag

2016). Current best practice is to reperfuse the injury as quickly as possible after ischemia

(Boag 2016, Stuart 2016). Reperfusion is achieved through primary percutaneous

coronary intervention (PPCI), which reestablishes blood flow to the area (Boag 2016).

Time is of the essence after onset of ischemia, After 90 minutes of occlusion, capillaries

begin to plug with thrombi and cells (Boag 2016). This is known as microvascular

obstruction, in which perfusion can’t be maintained even though blood flow has been

reestablished (Boag 2016). The accumulated cells also contribute to ROS production and

contribute to inflammation (Boag 2016).

27

While reperfusion is generally accepted to improve outcomes, it is not without

additional complications. Injury from reperfusion may be responsible for as much as 50%

of infarct size (Boag 2016, Nah et al. 2009). This is mainly due to “Lethal reperfusion

injury”, in which previously surviving cells die due to the changes in ROS, pH, and

mitochondria collapse (Boag et al. 2016, Nah et al. 2009, Przyklenk et al. 2012, Przyklenk

2014).

T cells can be activated by danger associated molecular patterns (DAMPs) (Boag

2016). T cells help limit infarct size and limit neutrophil and macrophage populations

(Boag 2016). T helper cells also produce IFN-γ and TNF-α and support macrophage

activity. Mice without lymphocytes saw reduced leukocytes in the infarct area, and

reduced injury from reperfusion (Boag 2016, Feng et al. 2015, Nah et al. 2009). Innate

immune cells (neutrophils, monocytes, etc) are thought to play a large role in regulating

inflammation (Boag 2016, Feng et al. 2015). Inadequate macrophage population can

prevent full scar tissue formation, potentially resulting in rupture (Stuart et al. 2016).

Clinically, high neutrophil counts have been linked to a greater risk of mortality (Boag

2016). A reduced lymphocyte population and increased neutrophil population after

reperfusion is a predictor for complications 3 years post-infarct (Boag 2016). The

increased risk associated with high neutrophil populations may be due to the increased

expression of MMPs. Excessive MMPs reduce scar tissue deposition and cross-linking,

resulting in a thinner, weaker scar, which allows for excessive left ventricle dilation

(Voorhees et al. 2015).

In MI without reperfusion, neutrophil population peaked at day 3, macrophages at

day 7 (Boag 2016). With reperfusion, the peaks for leukocytes remained the same

28

although overall cell counts were lower (Boag 2016). Neutrophil elastase (NE) is overly

active in chronic inflammation, in part because the α-proteinase inhibitor that would

counteract NE’s activity is itself deactivated by reactive oxidative species (Doring 1994).

Proper regulation of the inflammatory phase is crucial, as inadequate inflammation

can be just as detrimental as excessive inflammation. While inadequate or inappropriate

inflammation can lead to underdeveloped scar tissue, excessive inflammation however,

can lead to overly fibrotic tissue (Stuart et al. 2016). In these cases, scar tissue can extend

well past the injured area, affect mechanical and electrophysiological heart function. This

can lead to arrhythmia or cardiac arrest (Stuart et al. 2016).

Similarly, overactive and underactive leukocyte activity during inflammation can

lead to impaired healing and poor left ventricular remodeling (Stuart et al. 2016, Nah et

al. 2009). This poor remodeling can also result in arrhythmias due to poor

electrophysiological remodeling (Stuart et al. 2016). Therefore, even patients that survive

the initial infarct can be at risk for heart failure or cardiac arrest (Stuart et al. 2016). The

border region of the infarct zone is at the greatest risk for inflammation-induced

arrhythmias, which can affect the surrounding, healthy tissue (Stuart et al. 2016). Over

expression of TNF-α reduces cardiomyocytes’ ability to repolarize, directly affecting

potassium channels (Stuart et al. 2016). It also affects gap junction formation and

conduction velocity by limiting the promoter of the connexin 43 gene, which is responsible

for gap junction formation (Stuart et al. 2016). All of this can lead to arrhythmias.

Excessive fibrosis in the interstitial space of the border zone can also result in

arrhythmias, due to non-conductive collagen (Stuart et al. 2016). Therefore, induction of

an overly fibrotic state post-MI can be as detrimental as under-production of scar tissue.

29

It is also important to recognize the role of comorbidities in the inflammatory

response post-MI. Patients experiencing myocardial infarction frequently do so as a result

of coronary artery disease, and often with other high-inflammation pathologies, such as

obesity (Przyklenk 2015). This baseline level of chronic inflammation will clearly influence

the inflammation response in an injury such as MI, and this baseline level of chronic

inflammation is not adequately reflected in most animal models and therefore may

present some difficulty in predicting clinical effects.

Biochemical Inhibition of Inflammation

Numerous over-the-counter anti-inflammatory agents are used daily by millions of

people. One of the most common of these is ibuoprofen. Ibuprofen is a nonsteroidal

antiinflamatory drug (NSAID). All NSAIDs act on cyclooxygenase enzymes, which have

two forms: cyclooxygenase 1 and 2 (COX 1 and COX 2) (Amer et al. 2010, Ong et al.

2013, Kirkby et al. 2016). Due to its effects on the GI tract, many NSAIDs are now

designed to selectively block COX 1, which still enables the majority of the anti-

inflammatory effects (Kirkby et al. 2016). However, stable presence of COX 2 in the brain

and thymus require limited use of NSAIDs, including ibuprofen (Kirkby et al. 2016,

Patrono 2016).

Ibuprofen blocks the catalytic site from arachidonic acid via acetylation of a serine

residue near the binding site on COX 2, preventing prostaglandin formation (Amer et al.

2010). Prostaglandins are mediators of inflammation (Amer et al. 2010, Giroux et al.

2000). The large acetyl group prevents cleavage of arachidonic acid into PGG2 and

PGH2, which indirectly prevents platelet formation (Amer et al. 2010). This results in

reduced platelet aggregation and inhibits epithelial cell proliferation, both of which are

30

desirable for patients at risk for MI. This may be of special importance in situations in

which reperfusion does not occur until 90 minutes or longer post-MI, as previously

addressed.

COX 2 is expressed by cells involved in inflammation, including macrophages,

fibroblasts, or endothelial cells (Giroux et al. 2000). Various cytokines induce COX 2

expression, including interleukins (Giroux et al. 2000). Cyclooxygenases (COX) convert

arachidonic acid to prostaglandins (Giroux 2000). Therefore, synthesis of prostaglandins

is inhibited due to ibuprofen’s inhibition of COX 2 (Ricciotti 2011). Prostaglandins broadly

act on G-protein coupled receptors that could alter the receptor structure and function,

affecting cells’ responses to inflammation (Ricciotti et al. 2011). NSAIDs bind to and

deactivate one of the monomers of the COX 2 dimer (Ricciotti et al. 2011). Ibuprofen

inhibits COX 2 without restricting the NO pathway, which is required for wound healing

(Kirkby et al. 2016).

Inhibition of COX 2 has shown to have negative cardiovascular effects, specifically

increased risks of MI, hypertension, and reduced outcomes of congestive heart failure

(Amer et al. 2010, Kirkby et al. 2016). Because of this, NSAIDs are linked with increased

mortality after MI (Amer et al. 2010). However, a study by Leshnower et. al. showed that

administration of ibuprofen to rabbits and sheep after MI reperfusion did not result in any

significant changes in myocyte apoptosis or infarct size (Leshnower et al. 2006).

Additionally, ibuprofen is thought to induce signaling through NF-κβ pathway, which is

also involved in the inflammation process (Leshnower et al. 2006). Clearly, this shows

that ibuprofen could have a potential benefit in an overly inflammatory state, but may be

counterproductive to normal healing. While ibuprofen, along with other NSAIDs, has been

31

suggested to increase risk of subsequent MI after initial MI, ibuprofen is thought to be one

of the lower risk NSAIDs (Ong et al. 2013).

Naproxen (Aleve, Midol) is the only NSAID that has no risk of acute myocardial

infarction (AMI) associated with its use (Vargas-Lorenzo 2013). All other NSAIDs have

approximately the same small increased risk of AMI associated with daily use (Vargas-

Lorenzo 2013). However, this risk was analyzed from numerous studies that looked at

high-risk patient groups, which are far more likely to have an adverse event, independent

of NSAID use. Newer research looking specifically at ibuprofen has shown that rats

receiving ibuprofen after MI may have increased integrity in cell membranes, and reduced

ROS release (Patel et al. 2016). This suggests that ibuprofen may have some benefit as

a cardioprotective drug, or at least present less of a risk than other NSAIDs.

A study in which dogs with AMI were given a variety of treatments showed that

infarct expansion was attenuated by ibuprofen six week post-MI (Jugdutt et al. 2007).

NSAIDs can interfere with the healing and inflammatory processes, resulting in less

collagen deposition and net scar thinning (Jugdutt et al. 2007). This is of particular

concern clinically, where pathologies such as hypertension can lead to high intracardiac

pressures, making the infarcted tissue’s mechanical stability even more precarious.

Animals treated with ibuprofen only saw a decrease in mechanical strength of the infarct

area relative to sham, but not a significant decrease relative to infarct only animals

(Jugdutt et al. 2007). No change in distensibility, ejection fraction, or contraction strength

was reported relative to the control group, however a decrease in both collagen I and III

was noted (Jugdutt et al. 2007). Notably, the ratio of type I to type III collagen was

increased, despite this overall drop in collagen production (Jugdutt et al. 2007). These

32

results suggest that an NSAID, such as ibuprofen may be of some assistance in a

myocardial infarction treatment in which the mechanical integrity of the infarct area is

assured. Such would be the case in a patch application.

Use of Patches to Increase Mechanical Integrity Post-MI

Post-MI application of a biocompatible patch, with or without stem cells or growth

factors, is one of the most commonly researched treatments in cardiac bioengineering

(Lam et al. 2013, Wendel et al. 2014, Zhang et al. 2015, Tan et al. 2009, Serpooshan et

al. 2013). The premise of the research is that application of a patch to the infarct area

will increase wall thickness and provide mechanical support to the injured region (Zhang

et al 2015, Wan et al 2013, Stuart et al 2016, Wan 2017, Tan et al. 2009). By supporting

the damaged tissue, the wall stress on the remaining tissue is reduced. This prevents

scar bulging during systole as well as preventing excess mechanical strain on

surrounding tissue (Lam et al. 2013, Wendel et al. 2014, Zhang et al. 2015, Wan et al.

2017). Mechanical strain on the surrounding tissue has previously been described to

result in infarct expansion and increase risk of arrhythmia (Stuart et al 2016, D’Amore

2016). Some studies suggest that timely application of the patch can prevent changes in

ejection fraction and other measures of heart function as early as one week after infarct

(Wendel et al. 2014). Furthermore, by reducing the size of the infarct, patch treatments

have also been shown to increase contractility post- MI (Serpooshan et al. 2013).

Along with providing mechanical support, cardiac patch research also investigates

potential uses as a delivery vehicle for cells, growth factors, or medications (Wendel et al

2014, D’Amore et al. 2016, Tan et al. 2009). For this reason, natural and/or degradable

materials are preferentially investigated. However, degradation, particularly of synthetic

33

materials, has been cited as a potential risk for chronic inflammation in cardiac patch

applications (D’Amore et al. 2016).

In one study that looked at inflammatory effects, pigs underwent myocardial

infarction surgery and received with a multilaminate urinary bladder-derived ECM patch

(UBM) or expanded polytetrafluoroethylene (ePTFE) patch. Inflammation was reported

as increased in ePTFE compared to UBM, with UBM subjects displaying greater

myofibroblast recruitment and population (Robinson et al. 2005). While this indicates that

a naturally-derived collagen-based patch may result in a better prognosis for the patient,

it doesn’t fully explain how the inflammatory process is affected by a collagen-based

patch. In order to fully understand the implications of a cardiac patch therapy, it is crucial

to understand how these patches affect inflammation post-MI. These questions will be

addressed in Chapter 4.

When treating myocardial infarction, the current clinical goals are to limit scar

expansion, preserve contractility and encourage regeneration (Perea-Gil et al. 2015).

Current approaches involve use of a patch material, either with or without stem cells,

growth factors, or peptides (Perea-Gil et al. 2015, Tan et al. 2009). Collagen-based

patches have been included in numerous studies and have been shown to promote

angiogenesis and limit infarct expansion when administered within hours of MI (Perea-Gil

et al. 2015, Mewhort et al. 2014, Tan et al. 2009, Serpooshan et al. 2013). Importantly,

SIS patches have been used in murine infarct models and shown to increase ejection

fraction by as much as 35% post-MI (Toeg et al. 2013, Perea-Gil et al. 2015). In general,

performance changes are significant at least by 21 days post-infarct in murine models

(Toeg et al. 2013, Voorhees et al. 2015, Tan et al. 2009). However, in the literature

34

reviewed, significant performance (i.e. ejection fraction) changes were not noted at day 7

or earlier indicating that these changes occur primarily in early proliferation phase (Toeg

et al. 2013, Voorhees et al. 2015). This is logical, as it is only after inflammation that the

bulk of the scar tissue is formed and full integration of the patch material can take place.

Other murine models utilizing multiple patch models, showed that all tested

patches reduced infarct area size and increased wall thickness 8 weeks after infarct

(D’Aleno 2016). These findings suggest that the mechanical support provided by the

patch can have significant physiological improvements. Furthermore, these patch

materials have shown to retain any seeded cell populations or growth factors better than

synthetic counterparts, suggesting that collagen-based patches may be especially

beneficial in future tissue engineering applications (Perea-Gil et al. 2015, Mewhort et al.

2014).

While there is a great deal of research to support the mechanical efficacy of this

treatment, it is relatively unknown how patches directly affect the inflammatory process

(Zhang et al. 2015, Wan et al. 2017). What work that has been done has focused on

identifying risk of long-term chronic inflammation, and almost exclusively for synthetic

patches (D’Amore et al. 2016). This suggests a large gap in literature as it is apparent

from the literature that modifications to the inflammation process can greatly increase or

decrease the risk for arrhythmias, rupture, or other complications. It is also well

established that collagen-based patches provide excellent mechanical support,

consistently increasing left ventricular wall thickness and reducing infarct size in both

porcine and murine models (D’Amore et al. 2016, Zhang et al 2015, Perea-Gil et al. 2015,

Lister et al. 2016). Natural materials, including SIS are decellularized before use, using a

35

combination of chemical and mechanical treatments that are generally proprietary.

Despite these treatments, actual confirmation of removal of immunogeneic material is not

conducted for SIS (Lam et al. 2009, Wu et al. 2012, Shahabiour et al. 2016, Sutherland

et al. 2015). As there have not been many reports of rejection, it is a relatively safe

assumption that any remaining immunogeneic material is not enough to cause rejection.

However, even small amounts of such material can alter the inflammation response (Beck

et al. 2016, Sutherland et al. 2015). It is therefore prudent to investigate how these

patches, such as the commonly used porcine SIS patch, affect inflammation.

36

CHAPTER 2. MECHANICAL CHARACTERIZATION OF INTACT AND SUTURED HUMAN SKIN

Publication: Meier EM, Siddiqui A, Tepper DG, Lam MT.

Journal of biomechanical behaviour of biomedical materials- Submitted

Elizabeth Meier assisted with experimental design. She was responsible for tissue

processing, data collection and analysis, and manuscript preparation.

Introduction

Skin acts as a physical barrier between the internal body and the environment.

Part of this role includes maintenance of tissue integrity while facilitating and responding

to movement, which requires a range of viscoelastic material properties. Like many

tissues, skin’s mechanical behavior is dependent upon the arrangement and composition

of extracellular matrix (ECM). ECM composition of skin varies not only between species,

but based on age, gender, size, and location on the human body as well (Edwards et al.,

1995, Firooz et al. 2016, Zhu et al. 2015, Zollner et al. 2013). Due to this variation, it has

been difficult to fully characterize the mechanics of skin. The few studies that have been

performed with human skin in vitro were performed at very low strain rates in the range

of 2-20 mm/min (Blackmore et al. 2014, Grady et al., 2009, Miller et al., 2000). Even within

this narrow range of applied strains, large variation has been noted between properties,

with elastic moduli reportedly ranging from 5-30 N/mm2 and ultimate strains from 35-

115%, although some of this variation is attributed to differences in age of the samples

(Blackmore et al. 2014, Edwards et al., 1995).

Additionally, while the mechanical properties of sutures have been investigated, to

our knowledge, no study has investigated how the mechanics of sutured skin compare to

37

those of normal skin, beyond the microscale (Flynn et al., 2010, Vesentini et al., 2003,

Wang et al., 2014). Such information is critical for understanding how suturing changes

the biomechanics of skin, when suturing would be considered appropriate relative to the

numerous alternatives being studied in the field, and how suturing may affect other,

similar soft tissues that may not be as easily tested, such as the meniscus. Most

immediately, this information might be particularly important to patients who are athletes,

or otherwise committed to an active lifestyle, or for patients that require physical therapy

after a surgery or other injury. For these patients, it is highly important that the mechanical

effects of sutures on skin are known at a macroscopic level, especially if it leads to

increased risk of further injury. Currently, there is not information readily available about

how sutured skin behavior may differ based on the strain rate and orientation of the injury

relative to the direction of movement. Here, we investigated the tensile properties of

human skin that was injured, sutured and then mechanically stretched at low (2 mm/min)

and high strain (100 mm/min) rates. These loading rates were selected to encompass the

range of loading human skin would be expected to withstand during a range of common

movements, such as walking or stretching (Blackmore et al 2014). Our hypothesis is that

if skin is injured with a linear, full thickness injury, sutured and stretched 90º to and in-line

with the injury, then elastic modulus will increase relative to intact skin, due to the high

stiffness of the suture material. Furthermore, it is anticipated that fracture strength and

failure strain will decrease, due to the stress concentration around the skin at the bite

points. Additionally, we hypothesize that these differences will be more pronounced in the

high strain group, both between strain applications and within groups.

Materials and Methods

38

Human tissue procurement was obtained in accordance with the Wayne State

University Institutional Review Board Protocol #054514M1E. Skin samples were donated

from patients undergoing abdominoplasty at

Henry Ford Health System (Detroit, MI) with no

known diseases. Skin from 17 Caucasian

female patients aged 39-52 were utilized.

Patient age and sex were selected based on

sample availability and predicted variations due

to age and sex (Zollner et al. 2013). Male and

non-Caucasian skin samples were excluded

due to inadequate sample numbers and

potential mechanical differences between

demographic groups (Edwards et al., 1995,

Firooz et al. 2016, Zhu et al. 2015, Zollner et al.

2013). Patient data can be viewed in Table 1. All samples were taken from the abdomen,

although exact location and orientation of bulk skin samples relative to the patient were

unknown and therefore considered to be random. Tissue samples were received on ice

and subcutaneous fat was removed using surgical scissors. Skin samples for baseline

testing were cut to approximately 65 mm x 10 mm pieces, with three samples per patient

sample prepared to account for any variation or slippage during testing. Skin samples

that were to be sutured were cut to approximately 65 x 60 mm pieces. A scalpel was used

Table 1. Patient data for skin use.

39

to apply a 40 mm incision. Incisions were sutured using a simple, interrupted suture with

6-0 Ethicon Prolene® sutures placed every 5-7 mm, per reported optimal suture spacing

for the skin location and suture type (Vesentini et al. 2003). Orientation was defined as

the orientation of the injury, or “cut”, to the axis of applied tension. Sample orientation

would be either 90º to the injury or in-line with the injury line. Illustration of this definition

can be seen in Figure 6. Following suturing, all samples were stored in phosphate

buffered saline (PBS) at 4oC until testing.

Tensile testing was conducted on either an Instron 5943 (low strain) or Instron

5984 (high strain) with BlueHill 3 software (Instron, Norwood, MA). Tensile testing was

performed at two strain rates: 2 mm/min and 100 mm/min. Samples were installed in the

Instron as shown in Figure 7. Prior to testing, each sample was adjusted and measured

Figure 7. Installation of tissue sample in Instron. (A) Skin samples (red arrow) are

clamped into tensile device using serrated clamps (green arrow) and measured before

stretching. (B) Schematic of the suturing patterns (left – perpendicular orientation; right –

parallel orientation).

40

for consistency. Following installation, samples were submerged in a PBS bath during

testing to provide adequate tissue hydration. During testing analysis, failure mode in

samples with sutures was defined as the first instance of failure of any individual suture.

Similarly, intact skin was considered to have failed when a visible tear was able to

significantly affect the applied stress as indicated by the BlueHill software during testing.

Patient samples for the control, 90º, and in-line groups were n=7, n=6, and n=6,

respectively. Again, each sample number consisted of three individual tissue pieces from

the same donor, each tested separately to account for variations in location or installation

in the Instron. All were included in data analysis unless noticeable slippage at the grips

occurred, in which case the individual piece would not be counted toward the sample

average. Tensile mechanics at the same strain rate (ex. 2 mm/min or 100 mm/min) were

compared using one-way ANOVA with Tukey post-hoc with p < 0.05 for significance.

Results from the power analysis yielded a β of 86.4, based on the differences from the

elastic modulus calculations. This indicates that the sample size is adequate for this work.

All statistics were calculated using SPSS.

Stress relaxation testing was also conducted. Samples were prepared to the same

dimension and suture patterns as with tensile testing. Samples were placed in the Instron

5943 using serrated clamps and stretched to 80% of maximum strain at a rate of 2

mm/min. Maximum strain was determined from the tensile testing results at 2 mm/min.

Once 80% of maximum strain was achieved, the strain gauge length was held constant

for 5 min, with changes in applied stress recorded (Saulis et al., 2002). Five minutes has

been used as a hold length in skin mechanical testing previously, and thus is appropriate

for comparison between published results (Saulis et al. 2002, Blackmore et al. 2014).

41

Results were expressed as average percentage decrease in stress, rather than absolute

stress, to normalize the data. Recorded decreases in stress were averaged for each of

the three groups: intact (control), samples stretched 90º to injury, and samples stretched

in-line with injury. Percent changes in stress between groups were compared using one-

way ANOVA with Tukey post-hoc.

Results

Initial testing of intact skin samples at 2 mm/min, showed that human abdominal

skin has an average elastic modulus of 7.76 MPa. At 100 mm/min, measured modulus

significantly increased more than threefold to 29.1 MPa (p = 0.0001). However, these

results fit within the range of 5-30 MPa commonly cited as the range for adult human skin,

which illustrates the variation in reported data (Edwards et al., 1995, Blackstone et al.

2014). These two data sets were compared using a student’s t test to illustrate the range

in reported data.

Interestingly, the elastic modulus at higher strain rates reflects that of other

materials that consist primarily of collagen I, such as porcine small intestine submucosa

(SIS) (Roeder et al., 1999). This suggests that at higher strains, collagen I is the primary

component bearing the load, while other components, such as elastin, have a greater

impact on mechanical properties at lower strain rates. Such assessments can be based

on the high stiffness and nearly perfectly elastic behavior of collagen, which is significantly

higher than elastin or proteoglycans. This is also supported by the lack of plastic

deformation under 100 mm/min strain, whereas a small region of plastic deformation was

observed at 2 mm/min strain as shown in Figure 8. Samples with sutures had elastic

moduli almost double that of intact skin at low strains as shown in Figure 8 and Table 1.

42

This is likely due to a level of stress off-loading by the stronger, stiffer suture material,

protecting areas of skin from experiencing the full applied stress before the stress

concentration at the bite location overwhelmed the tissue. At higher strains, skin samples

stretched 90º to the injury had insignificant differences in elastic modulus measurements

43

as the control group (29.1 MPa), while recorded moduli for the samples stretched in-line

with the injury had a significant increase to approximately 39 MPa (p = 0.0001). This is

likely a result of the additive mechanical properties of the sutures, as well as an effect of

unbroken collagen filaments in the tensile axis. Fracture strength also saw a large

Figure 8. Tensile behavior of skin under low and high strain rates. Graph shows skin

under 2 mm/min strain (A) and 100 mm/min (B). Lower strain resulted in lower elastic

modulus (slope of the fitted line) and exhibited slight plastic deformation. Failure strains, or

maximum strain achieved before failure, were noted to be lower at higher strain rates.

44

increase between the strain rate groups, an observation similar to the elastic moduli.

Again, this suggests that at the higher strain rates, the mechanical behavior of collagen

is dominant.

Failure strain averaged 81.54% at 2 mm/min, which falls within the age-dependent

range of 35-115% strain before failure (Edwards et al., 1995). However, at 100 mm/mm,

failure strain was found to average only 44.5%. The lower failure strain may be due to

high strain rates’ limiting collagen fibers to straighten and align before being exposed to

higher levels of stress, resulting in failure at lower strains. While failure strain of the human

skin samples was found to be greater when stretched at 2 mm/min, the relationship

between the failure strains at each speed was found to be similar in the control (intact)

and 90º groups. Both groups saw a decrease in failure strain when the strain rate

increased from 2 mm/min to 100 mm/min. Samples stretched with the injury in-line with

the tensile axis, however, only displayed a decrease of approximately 23% with the

increase in strain rate. However, the failure strains were markedly lower in the in-line

group, with failure occurring at 46.8% strain at 2 mm/min and 36.2% strain at 100 mm/min.

Interestingly, the failure strains for both sutured groups at 100 mm/min are nearly

identical, with failure occurring around 35% strain, regardless of suture orientation (90º to

or in-line with the injury). Differences in failure strain at 2 mm/min were found to be

significant, between control and 90º to the injury (p = 0.0073), or in-line with the injury (p

= 0.001), and between the two suture groups (p = 0.0057). At 100 mm/min, none of the

differences between treatment groups were statistically significant. This suggests that the

generation of stress points created by the suturing process may result in early failure of

sutured skin at high strain rates. The comparatively higher forces during the high strain

45

rate may circumvent the viscous behavior of skin. This could lead to more rapid tearing

of the skin, thus resulting in undesired failure at lower strains. A summary of these results

can be found in Table 2.

Overall, the strain rate had a statistically significant impact on elastic modulus,

fracture strength and strain at failure. The elastic modulus was significantly impacted by

orientation, either 90º to the injury, or in-line with the injury, when stretched at 2 mm/min

(p=0.003 between control and 90º and p= 0.001 between control and in-line), however at

100 mm/min significance was only observed for skin stretched in-line with the injury (p=

0.001). Similarly for fracture strength, only samples stretched in-line with the injury had a

significant difference at both strain rates (p = 0.0127 at 2 mm/ min and p = 0.002 at 100

mm/min).

To test stress relaxation, samples were held under constant load and changes in

stress over time were recorded. Stress relaxation testing showed relatively consistent

relaxation proportional to the initial applied stress to achieve 80% strain as shown in

Figure 9. However, due to the differences in stiffness at lower strain rates, the net stress

relaxation varied between groups slightly, with samples stretched in-line with the injury

having the greatest net stress relaxation, but a slightly lower proportional relaxation (39%)

Table 2. Summary of tensile properties of skin samples without sutures, and with

perpendicular and parallel oriented sutures. * Indicates statistical significance within strain

rates (p <0.05).

46

than samples stretched 90º to the sutured injury (42%). However, none of the differences

were significant. These stress relaxation values are slightly greater than the results

published by Saulis et al., although this may be due to differences in sources of skin, as

well as age and gender variations, as the cited work investigated forearm skin flaps of

multiple ethnicities.

Discussion

When skin is sutured, a high-tensile strength material is used to bear tensile loads

applied to the skin (Wang et al., 2014). Suturing requires induction of micro and macro

tears in the surrounding tissue, creating major failure points as observed during tensile

testing (Flynn 2010). Additionally, these tears are subjected to forces averaging 0.5 N-

1.5 N required to pull the injury closed, further exacerbating the damage inflicted during

needle driving (Capek et al. 2012). Samples with sutured injury in-line with the tensile axis

had the highest average fracture strength of 4.86 MPa, compared with intact samples and

sutured samples stretched 90º to the injury, which had fracture strengths of 3.77 MPa and

3.54 MPa, respectively. Samples stretched in-line with the injury had the highest fracture

strength of all groups at 100 mm/min, with a fracture strength of 11.5 MPa. Additionally,

stretching sutured skin in-line with injury seemed to disrupt the viscous region of the

stress-strain curve, most notably at 2 mm/min. Such samples only reached 13.2% strain

before displaying elastic behavior. This is a noticeable decrease from the samples

stretched 90º to the injury (26.9%) and the intact skin groups (30.8%). Differences at 100

mm/min were much less noticeable, with all samples displaying strains in the range of 13-

17% before experiencing elastic behavior. This is of particular interest considering the

47

abdomen’s highly elastic behavior relative to other areas of skin, such as the head or

limbs (Blackstone et al. 2014).

While loading conditions in vivo are rarely uniaxial, uniaxial testing is appropriate

as a preliminary analysis of mechanical properties as it is more consistently and more

frequently performed on soft tissues (Blackstone et al. 2014, Zhu et al. 2015, Zollner et

al. 2013). In situations where loading is considered to be mainly, if not entirely, uniaxial,

other approaches have been taken to reduce stress concentration around each individual

bite. While simple, interrupted sutures are frequently used for closing superficial wounds,

a suture button with figure 8 suture pattern has been used with great success in reducing

stress concentration in patellar tendon repair (Otsubo 2016). This methodology did not

restrict range of motion and yielded a 98% success rate (Otsubo 2016). Additionally,

procedures such as anastomosis involve suturing of soft tissues that are very similar to

skin (i.e. adventitia) and these tissues experience uniaxial loading, especially during the

procedure (Roussis 2015).

Polypropylene (Prolene®) is a stiff and strong material, with an average elastic

modulus of 100 MPa (Greenwald et al., 1994; Chu et al., 1989). Prolene® has a tensile

strength of 40 GPa, and only minimally decreases after 6 weeks in vivo (Greenwald et

al., 1994). Given that the stiffness and strength of the suture material is significantly

greater than human skin, it is unsurprising that the sutures can increase stiffness and

decrease failure strain in skin. However, injury inflicted during suturing creates new failure

points. These data show that suturing can greatly affect skin’s mechanical properties, and

that the extent of the changes to the mechanical properties is highly dependent upon

suture orientation relative to the tensile axis as well as the applied rate of strain, with

48

greater strain rates resulting in increased stiffness of the sample. Additionally, these data

provide some interesting insights into the interaction between sutures, applied tension,

and the natural structural proteins found in skin. For example, sutures applied to an injury

in-line with applied tension may effectively off-load stress from viscous components, such

as glycosaminoglycans, as indicated by the low strains observed in samples before

engaging in purely elastic behavior. This also suggests that the high strength and stiffness

of sutures may dominate the local mechanics of a given tissue, but may be counteracted

by damage inflicted during the suturing process as indicated by the similarities in behavior

between the control group and samples stretched 90º to the injury. These data suggest

that we can partially accept our hypothesis that mechanical strength of injured skin with

sutures will increase relative to intact skin, as injured skin stretched in-line with the injury

displayed an increase in fracture strength. However, injured, sutured skin stretched 90ο

to the injury showed a slight (non-significant) decrease in fracture strength, although

elastic moduli were increased for both sutured groups relative to intact skin. It is also

important to note that the maximum strain for both sutured groups was much lower than

intact skin, which may be the limiting factor in addressing potential for further injury during

physical activity.

Conclusion and future work

By assessing comparative data under various strain rates and injury orientation

relative to movement axis, we aim to supply vital information for assessing potential risk

and complications for patients with injuries requiring sutures, particularly athletes

returning to training or physical therapy. By thoroughly understanding the effects of

sutures on skin’s biomechanical properties, the medical community can better advise

49

patients with injuries or procedures requiring sutures. Furthermore, the data presented

here can be extrapolated to predict behavior of other soft tissues used in tissue

engineering. In particular, decellularized dermis tissue, such as commercially available

AlloDerm ® are used constantly in plastic surgery procedures (Oh 2011, Shahabiour

2016, Jansen 2013). The decellularized dermis’ success has led to its investigated use in

abdominal repair surgeries, as well as providing the adventitia layer in vascular tissue

engineering (Jansen 2013). In both of these examples, a clear understanding of the failure

mechanism of suturing and tissue strength is crucial for patient outcome, yet these

scenarios are much more challenging to study using conventional technology. This study

can provide a reference for research such as this.

Figure 9. Stress relaxation results. A.) Initial and final stresses recorded for skin

samples without sutures, perpendicular sutures, and sutures in parallel. B.)

Percent decrease in applied stress after five minutes for each of the three groups.

No significant differences were noted between the groups.

50

CHAPTER 3. MECHANICAL STIMULATION DRIVES ADIPOSE STEM CELL DIFFERENTIATION TOWARDS A MENISCUS PHENOTYPE

Publication: Meier EM, Wu B, Siddiqui A, Tepper DG, Longaker MT, Lam MT.

Plast Reconstr Surg Glob Open. 2016 16;4(9):e864.

Elizabeth Meier was responsible for conducting experiments, including biochemical

stimulation and combined biochemical and mechanical stimulation. She compiled data

and assisted with statistical analysis and manuscript preparation.

Introduction

While the high tensile strength of sutures make suturing an attractive modality to

help facilitate repair, they are only practical in injuries capable of self-healing. In tissues

with limited ability to self-repair, such as the meniscus, other approaches must be

considered in order to facilitate wound repair. One such option is the use of stem cells to

create a new cell source to produce the correct proteins to develop a tissue with

comparable mechanical properties as native tissue. In tissues such as the meniscus,

which require mechanical loading to develop properly, it is imperative that mechanical

stimulation be included. Here, we seek to develop a protocol using a combined

mechanical and biochemical stimulus package to promote production of crucial structural

proteins.

One in six knee surgeries are due to meniscus related issues, and many meniscal

injuries are left untreated due to the lack of effective repair methods, causing long-term

damage and accelerated osteoarthritis (Brophy et al. 2012, Arendt et al. 2014, Hasan et

al. 2014, Rothrauff et al. 2016). Successful attempts to repair the meniscus with materials,

cells, or engineered tissues have been limited due to the difficulty of mimicking this

structurally and mechanically complex, heterogeneous tissue (Higashioka et al. 2014,

51

Rongen et al. 2014, Scotti et al. 2013, Yuan et al. 2014). Tissue donation from cadaver

sources are often ill-fitting with minimal improvement in mechanical function long-term

(Schuttler et al. 2014). Most meniscus engineering efforts are based on cell-populated

scaffolds but this approach has shown limited success (Schuttler et al. 2014, Bouyarmane

et al. 2014, Longo et al. 2013). Finding a viable cell source for meniscus tissue

engineering remains an issue. Stem cells are a promising option due to proven viability

in clinical trials (Pak et al. 2014).

Bone marrow mesenchymal stem cells (MSCs) have been a common cell type for

study in osteochondral applications, but are not practical for clinical use (Hatsushika et

al. 2014, Nerurkar et al. 2011, Okuno et al. 2014, Vangsness et al. 2014). Adipose-derived

stromal cells (ASCs) are abundant and easily obtainable, capable of being isolated from

fat tissue aspirated from the same patient. Our laboratory has previously shown

successful differentiation of ASCs down the chondrogenic pathway (Xu et al. 2007). ASCs

exhibit fibrogenic qualities, thus, we investigated the potential of ASCs to differentiate into

a mixed fibrogenic and chondrogenic phenotype. Furthermore, undifferentiated ASCs

have already been used in clinical trials, with some success (Pak et al. 2014)

Due to the knee’s dependence on mechanical loading in function and

development, mechanical strain may stimulate stem cells to differentiate into

mechanoresponsive cells (Arendt et al. 2014) Mechanical effects on bone marrow MSCs

has been studied extensively with inconclusive results (Byrne et al. 2008, Case et al.

2013, Kisiday et al. 2009). Here, we investigate the effects of mechanical stimulation on

the differentiation of human adipose-derived stromal cells (hASCs) towards a meniscus

phenotype. Biochemical factors were examined individually, then combined with uniaxial,

52

cyclic mechanical strain. Our results show that biochemical factors and mechanical strain

are capable of promoting fibrochrondrogenesis.

Materials and Methods

Stromal cell isolation and media

Human adipose-derived stromal cells (hASCs) were isolated from discarded fat

obtained from elective abdominoplasty and liposuction procedures. Specimens were

obtained with informed consent from patients in accordance with Stanford University,

Wayne State University, and the Henry Ford Health System human IRB guidelines. All

lipoaspiration and abdominoplasty procedures were performed by a board-certified

plastic surgeon. Patient age ranged from 18 to 65 years old, and adipose tissue for this

study were acquired from the flank or abdomen region. Cells from these regions were

selected based on their robust proliferation capability and morphological consistency.

Participating patients had no prior knowledge or evidence of ongoing systemic disease at

the time of operation. All specimens were immediately placed on ice and processed as

follows.

Lipoaspirates were washed twice in Betadine, followed by three rinses in

phosphate buffered saline (PBS). Whole fat was digested in 0.1% type II collagenase

diluted in Hank’s Buffered Salt Solution for 3 hours in a shaking water bath, assisted with

additional mechanical digestion via vigorous hand shaking every 30 min. The digest was

then neutralized by the addition of media with fetal bovine serum and centrifuged at 1000

rpm for 5 min. Supernatant with undigested fat was discarded and the remaining pellet

with stromal cells were resuspended in growth media (GM) and plated at 15 ml of original

tissue volume per 100 mm tissue culture dish.

53

The media formulations used are outlined in Table 3. Media components consisted

of Dulbecco’s modified Eagle’s medium with GlutaMax (DMEM, 10569, Gibco, Life

Technologies, Carlsbad, CA), fetal bovine serum (FBS, 26140079, Gibco), penicillin

streptomycin (PS, 15140163, Gibco), ITS+ universal culture supplemental premix (ITS+,

354352, BD Biosciences, Bedford, MA), L-Ascorbic acid 2-phosphate sesquimagnesium

salt hydrate (ASP, A8960, Sigma, St. Louis, MO), dexamethasone (Dex, D4902, Sigma),

and recombinant human transforming growth factor- β3 (TGF-β, 243-B3, R&D systems,

Minneapolis, MN).

Biochemical differentiation

Culture media for encouraging ASC differentiation into fibrocartilage-like cells was

determined from base fibrogenic and chondrogenic media formulations (Table 3). ASCs

were media differentiated for the customary 21 day culture typically used in chondrogenic

differentiation studies (Mandal et al. 2011). Cells of passage 1-4 were used for these

experiments and pooled. Fibrogenic media (FM) traditionally contains basal media and

fetal bovine serum to promote differentiation into fibroblasts. Chondrogenic media (CM)

is significantly more complex, with the addition of several chondrogenic growth factors-

insulin (ITS+), ascorbic acid (ASP), dexamethasone (Dex), and TGF-β. Two different

FM CM FM1 FM2

1% P/S 1% P/S 1% P/S 1% P/S

10% FBS 0% FBS 1% FBS 10% FBS

- ITS+ (1:1000) ITS+ (1:1000) ITS+ (1:1000)

- ASP (37.5 μg/mL) ASP (37.5 μg/mL) ASP (37.5 μg/mL)

- Dex (100nM) Dex (100nM) Dex (100nM)

- TGF-β (10 ng/mL) TGF-β (10 ng/mL) TGF-β (10 ng/mL)

Table 3. Summary of media components and concentrations.

Table 2

54

variations of combined fibrogenic and chondrogenic media were tested, composed of all

components of both medias with either 1% (FCM1) or 10% (FCM2) FBS. Serial media

differentiation was also assessed, by subjecting the cells to fibrogenic media followed by

chondrogenic media (FM-CM) switched halfway through the differentiation period (i.e. at

10 days), and vice versa (chondrogenic to fibrogenic media, CM-FM). Complete media

formulations and concentrations of supplements are listed in Table 3. Control media

consisted of minimal essential ingredients for cell survival, i.e. basal media and 1% FBS,

and was termed maintenance media (MM). All media contained 1% penicillin

streptomycin to prevent contamination. Media was changed in all groups every 3 days.

Mechanical strain device

A custom system was built for applying mechanical stimulation to the stromal cells

in the form of uniaxial, cyclic tension (Figure 10). Desired system build features included

an overall small device footprint, the ability to be completely housed in a traditional

incubator, and multiple sample chambers. The system was built to house four separate

culture chambers to enable simultaneous stimulation of multiple samples of cells of

varying parameters. The culture chamber consists of a disposable, commercially

available rectangular culture plate (Thermo Scientific Nunc Dishes, 267061, Waltham,

MA), which ensures sterility with each experiment at low cost. Silicone elastomer

substrates (polydimethylsiloxane or PDMS) were added into the culture chambers as a

stretchable cell culture surface as a means for applying mechanical strain. Inserts were

designed and machined out of acrylic to attach the silicone substrates to an actuating

stage. An actuating stage connects the culture substrates to the actuator, which applies

uniaxial, cyclic tensile strain to the cells. A bottom plate for holding the culture and

55

actuator setup together was fabricated out of Plexiglas, and allowed the entire system to

fit easily onto a single incubator shelf as a single unit.

Cell loading and mechanical strain regimes

The human adipose-derived stromal cells were subjected to several different strain

regimes to determine the ideal stimulation protocol for encouraging fibrochondrogenic

differentiation. To enable cell attachment to the non-cell adherent silicone elastomer

stretch surface, the substrate was functionalized with an attachment protein prior to

mechanical stimulation. Laminin (natural mouse laminin, 23017, Invitrogen, Life

Technologies, Carlsbad, CA) reconstituted in phosphate-buffered saline (PBS, 10010,

Gibco) was deposited onto the substrate at a concentration of 2 µg/cm2, enabling cell

adherence to the surface. Pooled ASCs between passages 1-4 were then plated onto the

polymer substrate. Following overnight attachment, the cells were loaded into the ethanol-

sterilized stretch device in preparation for mechanical strain application.

Mechanical stimulation parameters were varied one at a time to determine their

individual effects. Parameters varied were time of stretching, applied strain and frequency

of strain. First, cells were mechanically stretched for hourly durations of 1, 2, 3, 4, 5, and

6 hours, while parameters kept constant were strain at 10% and frequency at 1 Hz. At

each hour time point, cells were harvested for gene expression analysis. Based on results

from the biochemical differentiation (described in the Results section), ASCs were

stretched in fibrogenic and chondrogenic medias, separately. The second parameter

varied was strain. Strains of 5%, 10%, 15%, 20%, and 25% were applied to the ASCs for

3 h and 6 h, while frequency was kept constant at 1 Hz. Strain rates were chosen based

on the calculations below. Time points are based on results from the time varying

56

experiments. Lastly, frequency was varied from 0.5 Hz, 1.0 Hz to 1.5 Hz for 3 h and 6 h

under 10% strain. Unstretched ASCs served as controls for all experiments.

Gene expression

To determine fibrogenic and chondrogenic differentiation of the ASCs under the

various biochemical and mechanical stimulation parameters, quantitative reverse

transcription polymerase chain reaction (qRT-PCR) was performed. qRT-PCR was

carried out on each sample at the conclusion of each experiment: 21 d for media

differentiation, at the end of each time interval for mechanical strain (i.e. after each hour

time point), and after 3 h and 6 h after the various strain rates and frequencies were

applied. Briefly, total ribonucleic acid (RNA) was harvested from the cells using an

RNAeasy Mini Kit (Qiagen, Valencia, CA), the samples treated with DNAse I (Ambion,

Austin, TX), and the RNA amount quantified with a Qubit 2.0 fluorometer (Q32866, Life

Technologies). To convert total RNA to cDNA, reverse transcription (RT) was performed

by prepping samples using a Taqman Reverse Transcription Kit (Applied Biosystems,

Foster City, CA) and subsequently running the samples on a reverse transcription system

(GeneAmp PCR System 9700, Applied Biosystems). qRT-PCR was performed on an

Applied Biosystems Prism 7900HT Sequence Detection System. All reactions were

conducted in triplicate to verify technical robustness. All results were normalized to

GAPDH. Fibrogenic genes probed were collagen I (Col I) and versican (VCAN).

Chondrogenic genes probed were collagen II (Col II), Sox 9 and aggrecan (ACAN).

Collagen X (Col X) was also probed to mark cell hypertrophy, indicating the beginning

stages of endochondral ossification (Shen 2005).

Histological analysis

57

Hematoxylin & eosin (H&E) and Masson’s trichrome stains were performed on

cells following stretching for the strain rate and frequency varying studies. Immediately

following stretching, PDMS substrates with cells attached were removed from the stretch

device and fixed with 10% formalin. Following formalin fixation, samples to be stained

with Masson's trichrome were re-fixed in Bouin's solution for one hour at 56°C. Each

sample was then placed onto a glass slide. Slides were stained using hemotoxylin and

eosin and Masson's trichrome protocols, with the omission of xylene from both protocols.

During preliminary studies, it was determined that removal of xylene from the staining

protocols produced higher quality images and so was not included. Exclusion of xylene

did not adversely affect the stains. Samples were allowed to dry fully prior to imaging.

Statistics

All averages were calculated as mean values ± standard error. Statistical

significance of the media treatments alone was determined with one factor ANOVA tests

with (p < 0.05). All other experiments used a two-factor ANOVA to determine statistical

significance. Tukey post hoc analysis was performed to ascertain significance between

groups, with p-values set at < 0.05. Each stimulation trial was conducted a minimum of

58

three times (to ensure n=3 for each data

point) and utilized ASCs from eleven

different donors. Donor cells were pooled

so that ΔΔCt calculations were comparing

treated ASCs and control ASCs from the

same donor.

Results

Media Formulation

To determine the appropriate conditions

for promoting the fibrocartilage

phenotype, the human adipose stromal

cells were cultured in seven different

media formulations consisting of

variations of fibrogenic and chondrogenic

biochemical factors (media descriptions

are listed in the Methods section;

formulations listed in Table 3). Cell

morphology was observed throughout the

21 day culture period. Fibroblasts are

characterized by elongated, narrow cells

whereas chondrocytes are more rounded

in shape. In the different biochemical

Figure 11. Cell morphology under the

different media types. Black boxes indicate

the cells. In the FM media, cells were

elongated similar to normal fibroblast

morphology. In the FM-CM media, cells were

more rounded like chondrocytes, whereas in

the CM-FM media cells were more elongated

like fibroblasts. Scale bars = 100 µm.

59

combinations, the stromal cells exhibited morphologies consistent with fibroblasts,

chondrocytes and a combination thereof.

Maintenance media (MM) was formulated to contain the minimum essential

nutrients for cell survival, i.e. basal media with 1% FBS. This media served as the control,

however 1% FBS was not sufficient for cell survival for the entire 21 day culture period,

shown by cell degeneration and death (Figure 11). ASCs in fibrogenic media (FM)

consisting of basal media plus 10% FBS displayed morphology similar to fibroblasts.

Similarly, cells in chondrogenic media (CM) were rounded, resembling chondrocytes. In

the mixed fibrogenic and chondrogenic media (FCM1 and FCM2), cells were initially

rounded in appearance but gradually acquired a more elongated morphology similar to

fibroblasts, with few rounded cells evident in the culture. In FCM2 where the FBS content

was higher than in FCM1, cells were more fibroblast-like with very few rounded cells,

suggesting that cells mostly differentiated towards the fibrogenic phenotype under this

media due to the higher FBS amount. In the media switched from fibrogenic to

chondrogenic media (FM-CM), by 21 days cells were more rounded with some elongated,

fibroblastic-like cells intertwined. When media was switched from chondrogenic to

fibrogenic media (CM-FM), cells took on a more fibrogenic morphology by the end of the

culture period. In both cases of changing media halfway through the culture period, the

ASCs primarily acquired morphology similar to the latter media, indicating that the

dominating differentiation factor was the last media formulation to which the cells were

exposed. Specifically, CM-FM media produced more fibroblast-like cells and conversely

FM-CM media resulted in more chondrocyte-like cells. Although morphology is a useful

60

marker, a more quantitative measure is necessary for determining cell phenotype, hence

gene expression of the cells in the different medias was also investigated.

Biochemical formulations induce both fibrogenic and chondrogenic phenotypes

Expression of key fibrogenic and chondrogenic genes were probed to determine

the direction of ASC differentiation towards the fibrocartilage phenotype under the

different medias (Figure 12). Gene expression is presented relative to control,

undifferentiated ASCs. All medias resulted in expression to some degree of fibrogenic

and chondrogenic genes, even in pure fibrogenic or chondrogenic media. In fibrogenic

media (FM), the key fibrogenic gene collagen I increased in expression the most, followed

by a slight increase (100x) in the chondrogenic marker aggrecan. Chondrogenic media

(CM) resulted in a significant increase in the key chondrogenic gene collagen II

expression, and included increase in expression of chondrogenic Sox9 and fibrogenic

versican. Fibro-chondrogenic media mix 1 (FCM1, containing 1% FBS) and the

fibrogenic-to-chondrogenic media (FM-CM) only increased expression of collagen I, and

also showed increased levels of collagen X indicating cell hypertrophy. Fibro-

chondrogenic media mix 2 (FCM2, containing 10% FBS) only increased collagen X

expression. Chondrogenic-to-fibrogenic media (CM-FM) increased expression of

collagen I, Sox9, aggrecan, and collagen X. Overall, the significant difference in the

presence of the essential collagen II protein with combined increased expression of other

main chondrogenic and fibrogenic genes indicates that the chondrogenic media

61

formulation best promotes the

mixed fibrochondrogenic

phenotype. Hence,

chondrogenic media was used

in the mechanical stimulation

studies.

Mechanical strain device

For applying mechanical

strain with the desired

parameters, a device was

custom built as described in the

Methods section. The device is

able to apply up to 25% strain and up to 1.5 Hz as required by our experiments. The

device had no issues operating for the 6 hour time span of the experiments and could be

run for longer periods of time. The polymeric substrate was fabricated at a 1:15 polymer

base to curing agent ratio, rather than the traditional 1:10 ratio to ensure that the substrate

had enough strength to be able to withstand the tensile forces applied to it while still

remaining compliant. PDMS is an inherently hydrophobic material and thus does not

permit cell attachment. Thus, the substrates were pre-functionalized with the attachment

protein laminin to enable cell adhesion, allowing the cells to remain attached to the

substrate under the applied cyclic strain. For RNA extraction for gene expression

evaluation, cells were easily detached with a cell scraper. Cells could potentially be

Figure 12. ASCs exhibit increased chondrogen

ic and fibrogenic gene expression after culture in

the varying media combinations. Gene expression

represented relative to the control maintenance media

(MM). Gene upregulation of the major chondrocytic

marker collagen II was significant in the chondrogenic

media (CM). Error bars represent standard error. *

denotes statistical significance (p<0.05).

62

detached from the substrates and collected using trypsin. This would allow for their use

in tissue engineering a knee meniscus.

Mechanical Loading Results

The components that comprise the knee joint are highly dependent on mechanical

cues for proper development. It is believed that these mechanical cues participate in

differentiation of the local stem cells into the fibrocartilage cells of the knee, though the

precise mechanism remains to be elucidated. Thus, mechanical strain was investigated

as a potential stimulant of differentiation of adipose-derived stromal cells (ASCs), a

mesenchymal derived cell, down the fibrocartilaginous pathway. Mechanical stretch was

applied via uniaxial cyclic strain using the custom-built cell stretching device. FM was

used in all mechanical stimulation studies as the control media as it is the conventional

media ASCs are cultured in. CM was used because ASCs expressed the highest levels

of fibrochondrogenic genes in that media (Figure 12), and the greatest differentiation

towards fibrocartilage was desired.

Stretch duration effects

Cells were subjected to mechanical stimulation for hourly increments ranging

between 1 and 6 hours, while keeping constant strain at 10% and frequency at 1 Hz. The

cells were stretched in fibrogenic and chondrogenic media separately. Expression of

relevant fibrogenic and chondrogenic genes by cells subjected to stretch for varying time

durations in each fibrogenic and chondrogenic media is shown in Figure 13. In fibrogenic

media, collagen II expression was increased at early time points and versican increased

at 2, 3, 5, and 6 hours. Collagen X was increased at 3 and 4 hours, showing cell

63

hypertrophy. In contrast, in chondrogenic media collagen I and Sox9 expression were

distinctly increased across all time points. At 3 h of stretch, most fibrogenic and

chondrogenic genes were noticeably increased, suggesting that 3 h of stretch in CM

improved fibrochondrogenic differentiation the most.

Strain magnitude effects

ASC were subjected to mechanical strains of 5%, 10%, 15%, 20%, and 25% for 3

h

Figure 13. Mechanical stimulation for 1-6 hours promotes varying

expression of both chondrogenic and fibrogenic genes. Gene expression

represented as relative values compared to control ASCs. Strain was kept

constant at 10% and frequency at 1.0 Hz. Mechanical stimulation in chondrogenic

media mainly promoted increased expression of fibrogenic gene collagen I and

chondrogenic gene Sox 9. Statistical significance (p<0.05) was not observed for

these data.

64

and 6 h; in FM and CM; and at a frequency of 1 Hz. Results are shown in Figure 14. Gene

expression was clearly increased in comparison to the time varying studies, showing that

magnitude of strain had a greater effect on cell differentiation than time duration. At 5%

strain, fibrocartilage genes were more highly expressed after 6 h of stretch compared to

3 h. At 10% strain, 3 h of stretch resulted in higher fibrocartilage gene expression. At 15%

Figure 14. Phenotype expression varied with varying strain rates during mechanical

stimulation. Gene expression relative to ASC only controls under 5% -25% strain. Callouts

show 5% strain (A), 10% strain (B), 15% strain (C), 20% strain (D), and 25% (E). Frequency

was kept constant at 1.0 Hz. The data clearly shows a peak of genetic expression at 3 h of

stretch under 10% strain. Error bars represent standard error. * denotes statistical

significance (p<0.05).

65

strain, most genes increased in expression at nearly the same amounts. At 20% strain,

gene expression was slightly higher at 3 h compared to 6 h, with a noticeable increase in

cell hypertrophy at 6 h of stretch. At 25% strain, gene expressed was lower than at other

strain rates shown by a maximum relative expression of 107 as opposed to 1010. This

result indicates that the cells were exposed to too high of a strain rate, hindering their

differentiation potential. In the combined graph (Figure 14F), the greatest overall increase

in fibrocartilage gene expression was at 10% strain for 3 h, supporting the use of the

66

parameters of 10% strain and 3 h to promote the meniscal phenotype. Data at 15% strain

for 3 h and 25% strain for 6 h was statistically significant.

Histological analysis using H&E and Masson’s trichrome stains were used to

identify the presence of collagen at the protein level in the cells after stretching (Figure

Figure 15. Strain rates do not affect protein level collagen expression. Protein level

expression of collagen as seen in H&E (pink) and Masson’s Trichrome (blue) stains.

Control tissue with collagen present (i.e. skin) served as the positive control and is

displayed for comparison. Images taken at 40x reveal undetectable protein levels. This data

provides further evidence of the difficulty to achieve full translation of collagen genetic

expression into protein expression when differentiating from stem cells. H&E and Trichrome

stains display mostly cytoplasm and that cells survived at most strains rates, except cell

death can be observed after 6 h of mechanical stimulation at 20% and 25% strains.

67

15). In H&E, collagen is demarcated by dark pink and the cytoplasm by lighter pink. In the

trichrome stain, collagen is indicated by the color blue. For both time durations across all

strain rates, collagen could not be detected at the protein level. Only light pink indicating

cytoplasm was seen in the H&E stains, and no blue appeared in the trichrome stains. At

6 h of stretch and 20% and 25% strain, cell necrosis and death can be seen in the stain,

supporting the data showing a decrease in gene expression under these parameters,

likely due to overstretching. Positive control stains of skin samples are shown in Figure

15 to show that the lack of collagen staining in the tissue samples was not due to technical

issues.

Frequency effects

Cells were mechanically stimulated at varying strain frequencies of 0.5 Hz, 1.0 Hz

and 1.5 Hz at 10% strain for 3 h and 6 h in chondrogenic media (Figure 16). Relative to

control, undifferentiated ASCs, all fibrochondrogenic genes increased in expression when

strained at the applied frequency range. A strain frequency of 1.0 Hz promoted the highest

increase in gene expression, resulting in a statistically significant increase.

68

Histological analysis of the cells stretched at varying frequencies show

undetectable levels of collagen protein expression (Figure 17). Light pink stains in the

H&E samples indicate cytoplasm, with very little dark pink areas indicating possible

collagen protein content. Similarly, in the trichrome stains, blue stain indicating collagen

could not be seen. The cells appeared healthy in all frequency groups. In comparison,

cell death could be seen at higher strain rates (Fig. X). This difference suggests that the

cells were able to adapt to the range of frequencies used, whereas strains over 20%

overstretched the cells and caused cell death.

Figure 16. Mechanical stimulation frequency affects phenotype. Stimulation frequencies

of 0.5, 1.0 and 1.5 Hz were applied to the ASCs for 3 h and 6 h. Strain was kept constant at

10%. Gene expression was significantly increased under 1.0 Hz stimulation for both time

points, suggesting that 1.0 Hz is the ideal strain frequency to subject stem cells to for inducing

fibrochondrogenesis. Error bars represent standard error. * denotes statistical significance.

69

Overall, the ideal parameters to promote fibrochondrogenic differentiation of

adipose-derived stem cells is 3 hours of stretch at 10% strain and a frequency of 1.0 Hz

in chondrogenic media, as shown by the greatest increase in expression levels of

fibrogenic and chondrogenic genes.

Discussion

Multipotent stromal cells from adipose tissue were chosen due to their abundance,

non-controversial nature, and potential for ease of harvest and transplantation to the

same patient. Human ASCs can be easily obtained through minimally invasive liposuction

procedures, and even a small amount of adipose tissue can yield millions of stromal cells.

For these experiments, unwanted adipose tissue was donated from liposuction

procedures in the clinic. More importantly, phenotypically, ASCs are similar to fibroblasts

and have been reported to differentiate into chondrocytes (Xu et al. 2007). This

combination of phenotypic characteristics make the ASC an attractive candidate for

fibrocartilage differentiation.

Fibrogenic proteins collagen I and versican were used as fibrocartilage markers as

they are critical to the function of fibrogenic tissues. Collagen I (Col I) plays a key role in

a tissues’ ability to withstand tensile forces and is found in tendons, ligaments, and other

connective tissues that experience large tensile loads. Collagen I’s role in the meniscus

is thus to allow the tissue to withstand tensile loads exerted onto it. Versican (VCAN) is a

proteoglycan that plays a large role in adhesion of cells to the extracellular matrix, and is

present in large amounts in fibrocartilage (Naal et al. 2008).

70

Chondrogenic proteins collagen II, Sox9, aggrecan, and collagen X were used to

suggest chondrogenic phenotype because of their significant presence in the meniscus

and their key functional roles. Collagen II (Col I) is an especially important component in

the meniscus as it allows for the tissue to withstand tensile forces. Sox9 is a key

chondrogenic differentiation factor. Aggrecan (ACAN) an extracellular matrix

Figure 17. Stimulation frequency does not affect protein level collagen

expression. Protein level expression of collagen as identified by H&E (pink) and

Masson’s Trichrome (blue) stains. Control tissue with collagen present (i.e. skin)

served as the positive control and is displayed for comparison. Collagen protein was

not observable at 40x magnification, indicating the insignificant amounts of collagen

protein present in the differentiated stem cells, which is in clear contrast to the

noticeable increase in collagen gene expression found in the PCR analysis.

Images for both H&E and Trichrome stains display mostly cytoplasm, and that cells survived

well at all frequencies tested.

71

proteoglycan that also aids cartilage in bearing compressive loads. Lastly, collagen X

demarcates endochondral ossification.

Protein expression and presence after differentiation treatments is often included

to strengthen claims of full or partial stem cell differentiation. Although protein expression

was attempted, protein quantities were too low to be detected in mechanically-stimulated

trials. This is expected, as treatment times ranged from 1-6 hours, which is considered

too brief of a period given the amount of cells used for any significant protein production

in adult chondrocytes, let alone pre-differentiated or immature cells (Pelaez et al. 2009).

Continuation of culture after treatment was considered, but ultimately rejected as gene

expression changes were dependent upon mechanical stimulation. Therefore, the cells

would risk de-differentiation if allowed to continue to culture after stimulation and effects

of mechanical stimulation would be lost.

The combination of the aforementioned fibrogenic and chondrogenic proteins give

the knee meniscus its unique phenotype, allowing it to meet its functional demands. The

meniscus experiences tensile and compressive loads- collagen I, collagen II and

aggrecan allow the meniscus to bear these loads. The concentration of these proteins

vary within different regions of the meniscus in relation to concentrations of tensile,

compressive, hoop, and shear stresses, further contributing to the heterogeneity and

uniqueness of the fibrocartilage phenotype.

Adipose-derived stromal cells were treated in the various media formulations over

a period of 3 weeks to allow for progress towards chondrogenic differentiation. By

comparison, pre-fibrogenic differentiation occurs at a much faster rate, taking from a few

72

days to a week. Progression towards a differentiated state, or “pre-differentiation” was

evidenced by a changed morphology and enhanced relative gene expression by the cells.

For pre-chondrogenic differentiation, the ASCs would appear more rounded in shape,

indicative of chondrocytes. Under fibrogenic media, ASCs assumed a more elongated

shape with many new cellular extensions, resembling natural fibroblast morphology.

For the fibrogenic-to-chondrogenic (FM-CM) and chondrogenic-to-fibrogenic (CM-

FM) medias, the objective was to promote the mixed fibrochondrogenic phenotype. The

media was switched between the two media types halfway through the culture period, i.e.

10-11 days through the total 21 day culture, to maximize time exposure of the cells to

each media. Interestingly, the ASCs appeared to acquire characteristics of the latter

media, suggesting that maintaining stromal cell adaptations is dependent upon

continuous stimulation. Switching medias did not result in a mixed phenotype.

Expression of collagen I was seen in nearly all media groups that contained

fibrogenic media components. This may indicate that 1) adipose-derived stromal cells

have a tendency for fibrogenic differentiation given the correct fibrogenic media

conditions, 2) fibrogenesis may be the dominate differentiation direction given traditional

media components (since FM is the conventional media for culturing ASCs), and 3) the

presence of fetal bovine serum may be an adequate stimulus to encourage fibrogenesis-

like differentiation. No increase in collagen I expression was seen in the fibro-

chondrogenic media 2 (FCM2) compared to increased collagen I expression in the fibro-

chondrogenic media 1 (FCM1), indicating more fibrogenesis even though the serum level

was lower. This was possibly due to the fact that in FCM2, and under the conditions of

10% FBS and 3 weeks of culture, the cells began to overgrow and may have transformed

Figure 10. Mechanical stimulation device.

Image of the custom-build mechanical cell

stretching device.

73

into other, unknown phenotypes as commonly seen in cell culture overgrowth. Collagen

II expression, a chondrogenic marker, was understandably highest in the chondrogenic

media group (CM). No relative increase in collagen II expression levels was seen in the

FCM1, FCM2, FM-CM, and CM-FM groups, despite the presence of chondrogenic

factors. This effect is most likely due to the dominance of the fibrogenic components in

the medias.

Considering that chondrogenic media resulted in the highest relative expression of

both multiple fibrogenic and chondrogenic genes, this media was determined as the

optimal media formulation to obtain the mixed phenotype during the tensile strain testing.

Although Collagen I expression was minimal under Chondrogenic Media, cyclic, tensile

strain has been shown to increase Collagen I gene expression and encourage

fibrogenesis (Connelly et al. 2010, Kessler et al 2001). This suggests that using

chondrogenic media, which exhibited the highest collagen II expression while still

expressing the fibrogenic gene versican, has the greatest potential for obtaining a mixed

fibrocartilage phenotype when combined with cyclic tensile strain.

Following comparison of biochemical factors, tensile strain was added as a second

stimuli. Strains of 10% at 1Hz for 1-6 hours were initially applied, based both on our

preliminary calculations in Appendix A as well as support from other studies investigating

mechanical stimulation in connective tissue formation (Connelly et al. 2010, Kessler et al.

2001, Pelaez et al. 2009). In CM, ASCs subjected to mechanical stretch consistently

displayed higher levels of fibrogenic gene collagen I than in CM or FM alone, supporting

our hypothesis that the tensile strain would support expression of collagen I in the

presence of CM. At 3 hours of stretch in CM, the combination of fibrogenic and

74

chondrogenic gene expression was highest. At longer stretch duration, chondrogenic

gene expression decreased, while fibrogenic gene expression remained relatively high.

This is likely due to the dominance of mechanical stretch to produce more fibrogenic cells,

and at longer stretch times fibrogenesis gradually dominated over chondrogenesis

resulting in lower chondrogenic gene expression.

The three mechanical stimulation parameters investigated of stretching time

duration, strain rate and stimulation frequency showed different levels of effect on the

ASCs. The greatest increase in gene expression was seen in the frequency varying data,

followed by the strain rate data. Thus, frequency had the greatest effect on fibrocartilage

phenotype changes. Strain had the next greatest effect, and time duration had the least

relative effect.

There were no issues with removing cells from the stretching substrates for RNA

extraction, but a method will need to be determined in order to remove the cells as a cell

sheet for subsequent whole tissue meniscus tissue engineering. The cells cannot be

simply dissociated from the substrate as differentiated cells will not retain their phenotype

once dissociated and/or will simply apoptosis.

Chondrogenic differentiation is traditionally performed in a 3D culture, and usually

as a pellet culture. This is because chondrocytes require high cell density and close

proximity between cells. Hence, there is a level of difficulty in attempting to induce the

chondrogenic phenotype on a 2D surface, i.e. that of the stretch device substrate. There

is some evidence that chondrocytes and bone marrow mesenchymal stromal cells

(MSCs) are able to differentiate into cartilage under certain 2D conditions (Galle et al.

75

2010). While the factors of 2D culture and tensile stretch favor fibrogenesis, they do not

encourage chondrogenesis. Hence, chondrogenesis was mainly induced by the use of

chondrogenic media and by the chondrogenic growth factors within it. In this way, we

were able to still achieve a chondrogenic phenotype on a 2D surface.

Conclusion

We demonstrated for the first time, to our knowledge, the ability to encourage

fibrocartilage-like differentiation from human adipose-derived stromal cells using tensile

strain. Mechanical stretch along with biochemical factors promoted this mixed phenotype.

The challenge of producing the chondrogenic phenotype on a 2D surface, i.e. the stretch

substrate, was overcome by stretching in chondrogenic media. These fibrocartilage-like

cells provide a valuable stepping stone towards obtaining a fully mature, differentiated

fibrocartilage phenotype that could be used in tissue engineering applications down the

road.

76

CHAPTER 4. IMPACT OF CARDIAC PATCH AND IBUPROFEN ON INFLAMMATION FOLLOWING MYOCARDIAL INFARCTION

Publication: Meier EM, Wu B, Xu Z, Lam MT.

Tissue Engineering A- Submitted

Elizabeth Meier assisted with experimental design. She was responsible for

approximately half of all surgeries as well as post-operative care. She conducted PCR,

IF analysis, histology, and blood smear sample preparation. Bin Wu conducted the

remaining surgeries. Zhengfan Xu assisted with histology, immunofluorescence, and

sample preparation. Ashley Apil, Tiara Heard and Bijal Patel assisted with cell counts

and double blinding, respectively.

Introduction

So far we have determined that direct mechanical intervention can assist in wound

healing in certain tissues (ex. skin), but not in other tissues with limited capacity to heal

(ex. meniscus). Like the meniscus, heart tissue has limited capacity to regenerate, due to

extremely low cell replication. However, unlike the meniscus, heart tissue is highly

vascularized, resulting in a high capacity for inflammation and scar tissue formation.

These are the physiological events that occur in major cardiac tissue injuries, including

myocardial infarction. Myocardial infarction (MI), commonly known as heart attack, is a

sometimes deadly injury that can lead to heart failure and death.

One of the major complications of MI is cardiac remodeling and improper scar

tissue formation. One of the more popular research topics to address this remodeling is

to suture a biomaterial patch to the cardiac tissue to increase wall thickness and increase

cardiac function. Natural, decelluarized tissues, such as porcine small intestine

77

submucosa (SIS) are frequently used. These materials are attractive for MI treatments

because they have been shown to promote angiogenesis and retain delivered cells and

growth factors (Vanoos et al. 2016 Zhang et al. 2015, Mewhort et al. 2016 ).

However, it is largely unknown how addition of a collagen-based patch affects the

inflammatory process of the tissue. Collagen-based decellularized tissues have been

used in a wide variety of reconstruction surgeries (Jansen et al. 2013, Ibrahim et al. 2013),

with a wide range of inflammatory responses reported. Furthermore, many of these

materials incite a different inflammatory response between locations and applications in

the body, making extrapolation of results between applications ill-advised. For example,

AlloDerm, a collagen-based decellularized dermal layer, has minimal reaction when used

in breast reconstruction, but is quickly encapsulated and infiltrated by macrophages when

used in abdominal reconstruction (Broderick 2012, Ibrahim 2013). It stands to reason that

in the case of a myocardial infarction treatment, the physiological response would differ

from either of the two previous applications. While none of the documented responses to

SIS, AlloDerm, or any other similar decellularized tissue, have been severe enough to

warrant discontinuing use of the material, this variation suggests further study is needed

before any such material is used to treat severe injuries, like myocardial infarction.

To address this gap in knowledge, we propose to use a porcine small intestine

submucosa (SIS) patch on a rat infarct model. The SIS material has been used

successfully in infarct studies (Lam et al. 2013, Wendel et al. 2014, Zhang et al. 2015).

These studies have proven that a porcine SIS patch can improve cardiac mechanical

measures, such as ejection fraction, at least as well as other patch materials (Lam et al.

2013, Wendel et al. 2014, Zhang et al. 2015, Mewhort et al. 2016). Porcine SIS is

78

commercially available through Cormatrix, which has supplied the material for clinical

heart valve repair and pericardium replacement, among other uses (Gerdisch et al. 2014).

While it is an approved material with numerous studies conducted on its efficacy and

safety, its effects on the inflammatory response in an injury state, such as myocardial

infarction, are currently unknown. Due to the addition of a foreign material, we expect that

initial inflammation may be increased relative to infarct without any patch treatment. To

attenuate the increased inflammation, we expect that inclusion of an anti-inflammatory

agent, ibuprofen, may counteract this. To effectively evaluate how application of an SIS

patch may influence the inflammatory response post- MI, we plan to observe multiple

aspects of the inflammatory response in a rat MI model, with use of an SIS patch with

and without ibuprofen to determine how inflammation will be affected through day 7 post-

MI. Seven days was selected as the duration for the experiment as it has been suggested

as a rough timeline of the inflammatory response conclusion in a rat MI model (Turer

2011, Altara 2016 ).

Three subject groups will be used; infarct-only (control), infarct + patch, and infarct

+ patch +ibuprofen. Each group will be evaluated daily through seven days for gene

expression, protein expression, histology, and leukocyte counts. It is our expectation that

this analysis will present a clear view of the impact of a porcine SIS cardiac patch on the

inflammatory response after myocardial infarction.

Methods

Myocardial Infarction Model

79

Rats were selected for the infarct model due to their similar healing mechanisms

to humans. Myocardial infarction surgery was be performed in 200-225g female Sprague-

Dawley rats through left coronary artery ligation. Surgical procedures will be conducted

as outlined in the approved IACUC protocol A-03-06-14. Immediately prior to the surgery,

a baseline echocardiogram was performed, after which the animal would be randomly

assigned to a group. Animals were allotted into groups using a random number generator

(RANDBETWEEN function) in Microsoft Excel. A return value of “1” designated the animal

as infarct-only, “2” corresponded to patch only, and “3” received patch and supplemental

ibuprofen. Animals were anesthetized with isoflurane and intubated. The heart was

exposed via thoracotomy and the left descending coronary artery was occluded using a

6-0 suture slipknot with thread. The artery was occluded for 45 minutes, during which

moist sterile gauze was placed over the opening to limit dehydration. After the 45 minutes,

the slipknot was removed to allow for reperfusion and the patch applied (where

applicable). For designated animals, the SIS patch was visually aligned over the infarct

area and sutured into place using two 6-0 sutures. This process can be seen in Figure

18. After the patch is applied, the wound is closed and the animal will receive pain

medication and a bolus of saline to prevent dehydration. Analgesics were provided every

12 hours and MI confirmed via echocardiogram after 24 hours. Animals in the patch+

ibuprofen group will be given children’s ibuprofen daily via syringe feeding. Dosage of

supplemental ibuprofen was 30 mg/kg body weight/ day. Dosing was determined based

on pharmacokinetic analysis of ibuprofen in Sprague-Dawley rats (Fu 1991). Animals not

receiving ibuprofen were syringe fed with a placebo. Animals were sacrificed at days 1,

2, 3, 4 5, 6, or 7, with death confirmed via heart removal.

80

Gene Expression

PCR analysis was performed to assess the state of the tissue at each time point.

Key inflammatory markers MMP 1, MMP 2, and MMP 8 were examined. Collagen I will

be investigated for fibrosis and scar formation. TGF-β and TNF-α were reviewed for

information about the presence and type of

inflammation, along with IL-1RA. Neutrophil

presence and activity were indicated with

neutrophil elastase and neutrophil

chemotactic factor (CF) 2. Primers were

identified from Harvard Primer Bank and the

corresponding custom DNA oligos will be

ordered from Life Technologies. Primer

sequences are outlined in Table 4.

Following sacrifice, rat heart tissue will be crushed using a mortar and pestle.

Then, the steps outlined in the RNeasy minikit instruction manual (page 6-7) will be

performed to extract RNA. A Beckman Coulter Microfuge 18 Centrifuge will be used for

this process. Following the RNA extraction process, RNA will be stored in -80C or

immediately used in downstream steps. Before cDNA production can take place, we will

determine the concentration of RNA in each sample using a Qubit 2.0 Fluorometer and

reagents (Invitrogen/Life Technologies). To quantify RNA, 199 uL of Qubit RNA buffer

solution and 1 mL of Qubit RNA fluorescent reagent will be added to a 250 uL tubule.

Eight microliters will be extracted from the 200 uL mixture and 8 uL of the RNA solution

will be added. The quantity of 8 uL was selected due to the potential for low RNA

Figure 18. Placement of the SIS patch on

infarct area. After reperfusion, 6 mm patch

will be manually placed over the infarct and

held in place with sutures.

81

quantities in the infarct tissue. The tubule will be vortexed and then placed in the

Fluorometer for quantification. If RNA quantities are undetectable, the process will be

repeated with 20 uL solution removed from the 200 uL mixture and replaced with 20 uL

of RNA solution. Once the quantity of RNA is determined, preparations for cDNA will

begin.

82

Complementary DNA will be produced using Taqman Reverse Transcription

Reagents kit (Applied Biosystems Branchburg, NJ). The reagents will be mixed as

indicated in the kit instructions. Once the solution had been prepared, 6.525 uL will be

added to four 0.2 mL flat cap tubules (Fischer Scientific). RNA will be mixed with a

calculated amount of RNase-free water. Calculations are conducted so that equal

Gene Abbreviation Description Sequences

Collagen I Col I Expressed during fibrosis. GTTGCTGCTTGCAGTAACCTT AGGGCCAAGTCCAACTCCTT

MMP-1 MMP-1 Collagenase I, II, III MMP. Seen during remodeling and inflammation.

AAAATTACACGCCAGATTTGCC GGTGTGACATTACTCCAGAGTTG

MMP-2 MMP-2 Collagenase IV, gelatinase. Seen during remodeling and angiogenesis.

TACAGGATCATTGGCTACACACC GGTCACATCGCTCCAGACT

MMP-8 MMP-8 Neutrophil collagenase. TGCTCTTACTCCATGTGCAGA TCCAGGTAGTCCTGAACAGTTT

TGF-β TGF-β

Involved in inflammation and other processes. Produced by macrophages during inflammatory response.

AACTGCTTCCTGTATGGGGTC AAGGCGTCGTCAATGGACTC

TNF-α TNF-α Tumor necrosis factor, released by macrophages during inflammatory response.

CCTCTCTCTAATCAGCCCTCTG GAGGACCTGGGAGTAGATGAG

Neutrophil Elastase NE Serine protease released by neutrophils and macrophages during inflammation.

CTCGCGTGTCTTTTCCTCG GCCGACATGACGAAGTTGG

CD14 CD14 Monocyte, macrophage marker ACGCCAGAACCTTGTGAGC GCATGGATCTCCACCTCTACTG

Chemotactic Factor 2 CF2 Neutrophil chemotactic factor 2. Factor released during chemotaxis.

CCCACTCCCGGATTTGCTTC GTCTCGGTTAATGCTTCTGGTAA

IL-RA IL-RA Interleukin receptor antagonist 1. Produced by M2 macrophages.

GCGAGAACAGAAAGCAGGAC CCTTCGTCAGGCATATTGGT

Table 4. List of inflammatory marker genes and primer sequences.

83

amounts of RNA are included in each sample. Once the water is added to the RNA

solution, 3.475 uL of the resulting mixture will be added to each tube, totaling 10 uL. Once

the cDNA reagents are prepped, each reagent should be placed within a 2720

Thermocycler (Applied Biosystems). Once the program has finished its run, cDNA will be

stored at -20C for up to 24 hours, or used immediately in downstream applications.

PCR

PCR was conducted using StepOne Plus Real-Time PCR (Applied Biosystems).

The accompanying StepOne v2.2.2 software was used to set up the experiment. Each

gene and sample were assessed in triplicate and two controls used: negative control with

no cDNA, and cDNA from the MI-only tissue.

The plate design was executed in a MicroAmp Fast Optical 96-well reaction plate

from Applied Biosystems. Each well received 2.0 uL of the cDNA sample and 0.4 uL of

both the forward and reverse primers. Negative controls received 2.0 uL of RNase-free

water in lieu of cDNA. All wells received 7.6 uL of RNase-free water, for a total of 10 uL

of solutions. The plate was then covered with aluminum foil to block light as 10 uL Power

SYBR Green PCR Master Mix (Applied Biosystems) was added to each well. Once the

plate was loaded and cycles completed, gene expression fold-increase was calculated

from CT values provided by the software.

Differential Quick Stain

In order to determine the proportion of inflammatory cells within each time point

and group, a Differential Quik (Diff-Quik) blood smear stain was performed. This stain is

commonly used in immunology and inflammation studies to assess different inflammatory

responses (Van Hout et al. 2015, Geissman, et al. 2010, Godwin et al. 2016). Coronary

84

and ventricular blood was extracted from each tissue sample using a syringe immediately

after sacrifice. Blood was dropped onto labeled slides and spread using a spreader slide.

Each slide was fixed and stained using a

three step process using PolyScience’s

Differential-Quik Stain (Warrington, PA).

Once the stains set, groups of one

hundred cells were counted and

neutrophils and monocytes were tallied.

Neutrophils are signified by dark blue

nucleus and granules with pink cytoplasm.

Monocytes appear with lobulated, violet

nuclei with light blue cytoplasm, an

example is shown in Figure 19. Following staining, each slide was relabeled by another

researcher blinded to the study to prevent bias during data collection and analysis. In

order to ascertain quantitative differences, each smear was viewed at 6 different

locations, selected at random by two double blinded researchers with no relation to the

study. The leukocytes were tallied per approximately 100 cells, with the exact cell count

being included. The tallies at each location were averaged per sample, with three samples

being assessed per each treatment group at each time point, for six points for blood

smears from animals receiving infarct only, infarct + patch, and infarct with patch and

ibuprofen groups at days 1-7 post-infarct. Six sham animal smears were included as an

additional control.

Figure 19. Diff-Quik stain of example blood

smear. This smear shows a neutrophil and a

lymphocyte in the 100-cell count of the smear.

85

Histology and Immunofluorescence

Samples for histological examination will be excised and frozen at -80C in OCT for

a minimum of 24 hours. After samples have been thoroughly frozen, samples will be sliced

in 6 um sections using a Cryotome. Section thickness was determined during preliminary

staining to determine optimal thickness for staining visibility. Section slicing will be done

starting with the apex of the heart, near the infarct area, moving upwards along the heart

through the infarct zone. Each slice will be numbered, with every other slice being allotted

to either histology stain or immunofluorescence for a total of thirty slices equidistant

throughout the heart. Ten slices will be allotted for IF, while the rest will be allotted for

picrosirius red, and hematoxylin and eosin. This way, each stain will have an equal

representation of the infarct area while minimizing the number of animals needed. By

analyzing multiple slices throughout the infarct region, we will be able to evaluate the

distribution of key inflammatory cells throughout.

Histology

Histological evaluation provides qualitative evidence of tissue damage and infarct

size. Hemotoxylin and Eosin (H&E) and Picrosirius Red was used to delineate the infarct

area, healthy muscle, deposited collagen and the SIS patch, where applicable. H&E

staining was performed with 6 μm sections from two animals to confirm qualitative

evaluation. The same size slices was used for Picrosirius Red evaluation. Tissue samples

stained with Picrosirius Red were also observed under polarized light to aid in collagen

deposition and maturity analysis.

Immunofluorescence

86

Immunofluorescence (IF) staining was conducted to identify CD45AB, and was

used to assess the presence, location, and density of inflammatory cells (leukocytes) in

the infarct area. To separate collagen’s autofluorescence from the CD45AB signal, the

slides were imaged prior to IF staining and assessed for brightness using ImageJ (U.S.

National Institutes of Health, Bethesda, MD). The samples were then assessed using

ImageJ after staining to determine the difference in brightness, and therefore the

presence of leukocytes. The quantitative differences between the before and after images

were used to determine the relative presence of leukocytes between time points and

treatment groups. This was conducted for nine slices within each sample, with the same

corresponding slices evaluated between the groups. For example, if the tenth slice was

analyzed for one sample, it was analyzed for the remaining samples. Each slice was

imaged at a minimum of three locations to account for local variability. All brightness

measurements were included in the analysis.

Tensile Testing

It is well established that healed wounds exhibit different mechanical behavior than

that of native, intact tissues. Of particular note is the presence of scar tissue, which

increases the elastic modulus of the tissue and can alter the anisotropy of the tissue.

However, thick, intact scar tissue is not fully formed during the inflammation process and

therefore little is known about the immediate changes in mechanical properties following

an infarct. To evaluate the mechanical changes in the heart post-MI and how a collagen-

based SIS patch may affect these changes, tensile testing was conducted on all day 7

tissue samples. Tensile testing was conducted on a Cell Scale UStretch uniaxial

mechanical tester (CellScale, Ontario, Canada). Samples were stretched at a rate of 2

87

mm/min with data collected five times per second. Three tissue sections were procured

from each heart sample with three hearts used for each treatment group plus a sham

group. SIS patches were gently removed with forceps prior to tensile testing for applicable

samples. Elastic modulus, ultimate tensile strength and failure strength were calculated

and compared using one-way ANOVA with Tukey post-hoc tests.

Statistics

All averages will be calculated as mean values ± standard error. Statistical

significance for PCR was determined with two factor ANOVA with an alpha value of 0.05.

Tukey post hoc analysis was performed to ascertain significance within groups, with p-

values set at 0.05. Ad hoc power analysis was included to ensure sample size accuracy,

with a post-data power analysis yielding β= 0.76, which is close to the 0.80 used to

calculate the sample size. The same statistics were used for the leukocyte counts in the

blood smears. Brightness, or leukocyte presence in the tissue, was also assessed using

a two-factor ANOVA with the same parameters. Elastic modulus, ultimate tensile strength

and failure strength from tensile testing were calculated and compared using one-way

ANOVA with Tukey post-hoc tests. Sample sizes for quantitative measurements were

determined completing a power analysis. Initial sample sizes were estimated based on

published data from similar studies that confirmed the sensitivity of Differential Quik

staining (Selvi et al. 2000, Silverman 1995) and cell and gene expression in inflammatory

studies (van Hout 2015, Vuohelainen 2014). Histology, and polarized light microscopy

were evaluated qualitatively. All statistical design and analysis were completed with the

assistance of Wayne State University’s Research Design and Analysis (RDA) consulting

unit.

88

Results

Expression of Inflammatory Genes

PCR results were compiled and organized into groups of related inflammation

factors. Matrix metalloprotease (MMP) expression is shown in Figure 20A-D. MMP

expression overall showed minimal difference between the groups, with the exception of

significant down-regulation of MMP-2 at day 7 relative to infarct-only (Figure 20). As seen

in Figures 20E-H, neutrophil activity is overall elevated in the first 5-6 days post-injury

relative to the infarct-only group. However, by day 7, expression of all neutrophil markers

was below that of infarct-only. This, along with the elevated MMP expression during this

time, suggests that neutrophil activity, and potentially population, is elevated at

approximately days 3 – 6 in animals receiving the patch, relative to animals that received

only the infarct. The overall lack of difference in expression between the patch groups

suggests that the ibuprofen, either through mechanism or dosage, does not significantly

affect neutrophil gene expression post-infarct.

Figure 20I-L shows increased expression of key inflammatory markers in the patch

treatment groups at days 2-3 and at days 5-6 relative to the other time points. These

markers are expressed by, but not exclusively by, macrophages and thus only limited

conclusions of macrophage activity can be drawn. That said, these data suggests that

gene expression of these key genes can somewhat effectively “track” macrophage

presence, while providing information about the inflammatory process, as the gene

89

90

expression neatly follows the timeline of macrophage activity in inflammation outlined in

literature (Altara et al. 2016). This is particularly apparent when reviewing IL-RA

expression, which has clear peaks at expected macrophage infiltration times. Specifically,

the expression peaks during the first few days of inflammation, when the local, pro-

inflammatory macrophages (M1) are most often active, and then peaks again towards the

end of inflammation when the population of circulating, anti-inflammatory macrophages

(M2) is highest. Figure 20M-O supports the increased macrophage expression, as

collagen I expression is lower in the later stages of inflammation in both treatment groups

relative to the infarct-only group. Interestingly, during the middle days of inflammation,

collagen I expression is increased relative to the infarct-only group. This suggests that

overall collagen I expression is not lower between groups, but rather may be expressed

on an accelerated timeline in groups receiving the patch treatment after infarct. As shown

in Figure 20M, the expression of collagen I is related closely to the expression of TGF-β,

which is expressed by macrophages as well as other cell types in the area. Clearly, these

Figure 20. PCR results plotted over time relative to infarct only. Dotted lines indicate infarct

only expression. All significant results were relative to infarct only, indicated by * (P < 0.05).

Genes were grouped based on physiological relevance. MMPs 1, 2, and 8 are shown in A-D.

MMPs showed a clear decline in expression over time. Neutrophil genes CD14, neutrophil

elastase and CF 2 are shown in E-H. Neutrophil markers have a net decrease in expression

over time, and expression levels remain consistently lower than infarct only expression

levels at the end of day 7. Macrophage markers are shown in I-L, with TGF-β and collagen I

in M-O. Macrophage expression shows higher expression around day 2-3 and 5-6 relative to

the other time points. Error bars have been removed from combined graphs (A, E, I, M) for

clarity. I + P = patch only; I + P + IB = patch and ibuprofen.

91

data cannot give a complete picture of the differences in the inflammatory process

between the groups, however it can provide valuable evidence when combined with other

data, as described next.

Evaluation of Leukocyte Populations

Analysis of the neutrophil and macrophage cell counts was conducted using a two

factor ANOVA model in SPSS. These results are outlined in Table 5. The change in

neutrophil counts over time was significant for all groups (p = 0.004). Macrophage counts

also varied significantly over the 7 day period for all groups (p = 0.008). Significant

decreases in neutrophil counts were noted for all groups between days 2 and 6 (p = 0.047)

and days 2 and 7 (p = 0.014). Similarly, macrophage counts yielded significant differences

between early and later timepoints throughout the experiments for all groups. Significant

differences were noted between days 1 and 6 (p= 0.002), 2 and 6 (p= 0.022), 3 and 6 (p=

0.001), 4 and 5 (p= 0.049), and 4 and 6 (p= 0.008). These differences are likely due to

the low macrophage counts at day 4 and the higher counts at day 6. These results are

similar to other inflammation studies that show the greatest macrophage population

towards the end of the inflammatory period, generally around day 7 (Altara et al. 2016,

Stuart et al. 2016, Giroux et al. 2000, Fu et al. 1991). Therefore, these results do not

suggest significant ill effects from the treatments. Differences in neutrophil counts

between groups were significant at days 4 (p = 0.05) and 6 (p = 0.01), due to the higher

neutrophil presence in the patch + ibuprofen group. Differences in macrophage counts

were significant between groups at day 2 (p = 0.004), with the patch + ibuprofen group

having lower population than the other two groups. As macrophage populations remained

low in the group receiving ibuprofen, it may indicate that the ibuprofen slowed the

92

inflammation process slightly, delaying the arrival of circulating macrophages and limiting

the influx of cells. Also significant were the counts at day 4 (p = 0.03), this time due to the

lower macrophage population in the infarct-only group. This may be due to a heightened

response to the patch treatment, resulting in a greater population of macrophages

traveling to the injury area. An example of the smears can be seen in Figure 19.

93

Ta

ble

5. M

ean

s a

nd

sta

nda

rd d

evia

tion

s fo

r ne

utro

ph

il and

ma

cro

ph

age

co

unts

. No

sig

nific

an

t diffe

ren

ces w

ere

no

ted b

etw

ee

n

gro

up

s. S

ign

ifica

nt (p

< 0

.05

) diffe

ren

ce

s b

etw

ee

n d

ays w

ithin

trea

tme

nt g

rou

ps a

re d

esig

na

ted

by m

atc

hin

g s

ym

bo

ls. IB

= Ib

up

rofe

n.

94

Leukocyte Location and Density

While analysis the of the Quik Diff stains yielded useful information about the cell

populations, the cell counts do not provide information about the leukocytes’ locations

within the infarcted area. Tissue samples from day 7 stained with CD45AB antibodies

were viewed under fluorescent light and cell location and density observed (Figure 21).

The CD45AB antibody exclusively identifies monocytes and neutrophils in rats, although

it does not allow for identification between the two cell types (van den Berg 2001). It was

immediately apparent that leukocytes respond to the patch, as a high density of

leukocytes are located on the periphery of the patch in both the infarct + patch and infarct

+ patch + ibuprofen groups (Figure 21). However, there was no clear area of high

leukocyte accumulation in the infarct only tissue. It is highly likely that the majority of the

cells seen in these images are macrophages, due to the low neutrophil gene expression

(Fig. 20) and low neutrophil counts (Table 5) at day 7. To evaluate the differences in

leukocyte population within the tissue, images were analyzed before and after fluorescent

labeling of CD45AB antibody, with the resulting average difference in Table 6. The

average difference is the average readout of brightness due to the antibody labeling and

was performed to reduce readout error due to collagen and other background effects.

This provides a more direct comparison between groups, as images had varying levels

of background signal, which can influence the difference in the brightness readouts before

and after fluorescent labeling. The data shown in Table 6 overall shows a strong

correlation with the cell counts (Table 5). Overall, both patch groups generally had higher

brightness values, or higher leukocyte populations, within the tissue, than the control

tissue from the corresponding day, indicating that the addition of the patch resulted in

greater migration of leukocytes. The values at Day 1 seemed to display the most variation

95

between the cell counts and the ImageJ analysis, likely due to the sensitivity of the

inflammation process in the first 24 hours of the inflammatory process. The surgical

process allowed for variations of up to 6 hours from infarct until sacrifice for 1 day animals,

which may account for the variability in the leukocyte counts. Leukocytes, particularly

neutrophils, are activated within the first few hours of injury, and therefore a few hours

could make a large impact on population counts (Feng et al. 2015). However, after day

1, results were consistent and leukocyte population in the tissue appeared to follow the

same population flux as the cell counts, with an overall peak in cell population at Day 5.

These data combined suggest that the patch treatments do not necessarily prolong the

96

inflammatory response relative to infarct-only, but may result in an elevated inflammatory

response that is resolved quickly.

Figure 21. Immunofluorescent images showing DAPI and CD45AB (leukocyte) stains. A

higher density of cells can be seen around the patches in the patch treatment images. Cells

are more evenly dispersed in the infarct only samples. I = infarct; P = patch, C = collagen.

Scale bar = 200 μm.

Table 6. Summary of brightness difference between groups. Brightness indicates average

leukocyte presence. Overall, increases in leukocytes were seen throughout the 7 days with small

peaks around day 3-4 and 6-7.

97

Collagen Deposition in the Infarct Zone

Tissue samples from day 7 show the beginning of deposited collagen in all groups.

However, the samples receiving only the SIS patch and no ibuprofen after infarct had a

much more defined collagen layer than either of the other two groups. Of note is the

collagen surrounding the patch itself, which may be indicative of early encapsulation. SIS

implants been known to result in encapsulation in other uses (Jansen et al. 2013, Ibrahim

et al. 2013). The collagen

deposition on the cardiac

surface, however,

appears to be similar in

thickness and volume

between the two patch

groups. Interestingly, the

collagen in the infarct

only group appears to be

thinner, than in the

treatment groups (Fig.

22).

Collagen Maturity

Analysis

Figure 22. H&E and Picrosirius Red stained tissue samples.

“I” indicates the infarct area.. I = infarct; P = patch; C =

collagen. Scale bar = 1000 μm.

98

Picrosirius red stained slides were viewed using a polarized filter in a Zeiss

Axiovert 200. Under polarized light, mature collagen fibers appear red or orange while

thinner, less mature fibers appear yellow or green. Generally, infarct models show little

immature collagen fibers within the first week, but up to 95% mature collagen after 5

weeks (Wan et al. 2013, Rich et al. 2005). Figure 23 shows samples from infarct only,

infarct + patch, and infarct + patch + ibuprofen. All three groups showed some signs of

new collagen deposition: thin, immature green fibers in the infarct area. The most

interesting difference between the three groups was not in the infarct zone itself, but in

the collagen surrounding the patch in the treatment groups. As noted in the histological

analysis, the infarct + patch group had a much larger layer of collagen surrounding the

patch than the infarct + patch + ibuprofen group. The collagen also appeared to be more

mature than the collagen in the infarct zone, as indicated by the orange color. Also of note

was that the infarct + patch + ibuprofen group appeared to have more mature collagen in

both the infarct zone and in the area between the patch and the cardiac surface. As the

increased collagen is consistent between the two patch groups, it is likely that the patch

increases local inflammatory response. As there was consistently more collagen in the

patch only group relative to the patch + ibuprofen group, there is a possibility that the

99

ibuprofen was responsible for the difference in new

collagen deposition. Collagen I expression as shown in

Figure 20 O. supports these observations in that

collagen I expression was elevated in the groups

receiving the patch relative to the infarct only group, yet

expression in the patch + ibuprofen group was lower

than the patch only group, particularly during days 4

and 5. While this does not directly nor completely

explain the differences in the polarized light images, it

does support the overall observation that inflammation,

including the end-inflammatory state collagen

deposition, is increased with the patch use.

Tensile Mechanics

Tensile analysis following a seven-day period

showed relatively little differences between infarct only

animals and animals receiving the patch treatment.

However, elastic moduli were significantly increased in

animals experiencing myocardial infarction compared

to sham animals (Table 8; Fig. 24). While differences

in ultimate tensile strength were not determined to be

statistically significant, the data suggest that cardiac

tissue is weakened seven days post-MI relative to

uninjured tissue. Likewise, while statistically insignificant, a similar trend can be seen with

Figure 23. Images show faint

collagen presence in infarct

area. Thicker red and orange

patches can be observed

around the patches. I = infarct;

P = patch; C = collagen. Scale

bar = 1000 μm.

100

fracture strength, with fracture strength decreasing in samples with infarct (Fig. 24).

Interestingly, the greatest decrease in fracture strength occurred in groups that received

the patch and supplemental ibuprofen treatment after infarct. This difference was found

to be significant between the infarct + patch + ibuprofen group and the other treatment

and sham groups. Again, while not all tensile results were statistically significant, these

data do suggest that the tissue turnover and remodeling may be exacerbated or increased

in animals receiving the patch and supplemental ibuprofen after infarct. One explanation

could be that the increased exposed collagen from the patch results in an increase in

MMPs and other proteases, leading to increased removal of damaged tissue. This is also

101

supported by the slight increase in MMP-1 and MMP-2 during early inflammation as noted

in Figure 20.

Figure 24. Raw data graphs showing samples of data output from tensile loading.

All groups showed a similar toe region, indicating that overall composition is not

significantly different. IB = ibuprofen.

Elastic Modulus (kPa)

Ultimate Tensile Strength (kPa)

Fracture Strength (kPa)

Control 89.3 ± 73.2 148 ± 36.8 148 ± 19.1

Infarct Only 102 ± 42.0 * 143 ± 33.7 * 140 ± 18.3 *

Infarct + Patch 108 ± 88.1 * 134 ± 39.0 * 145 ± 28.0 *

Infarct + Patch + IB 111 ± 0.0450 * 135 ± 40.3 * 111 ± 69.7 *†

Table 7. Means and standard deviations for tensile measurements. The sham groups were

significantly different from the treatment groups (*). The ultimate tensile strength for the

infarct + patch + IB group was also significantly lower than the other groups (†). IB =

Ibuprofen.

102

Discussion

The function of inflammation is twofold: to clear the injured area of cell and tissue

debris, and to produce the biochemical progenitors of fibrosis and angiogenesis (Jugdutt

et al. 2007, Altara et al. 2016, Ricciotti et al. 2008). Unresolved or poorly managed

inflammation can lead to inadequate or excessive scar tissue deposition, which can have

devastating effects (Jugdutt et al. 2007, Bonaventura et al. 2016). Similarly, overactive

and underactive leukocyte activity during inflammation can lead to impaired healing and

poor left ventricular remodeling, resulting in arrhythmias and/or increased risk of heart

failure or cardiac arrest (Wan et al. 2016, Stuart et al. 2016, Giroux et al. 2000). The

border region of the infarct zone is at the greatest risk for inflammation-induced

arrhythmias, in which newly formed, non-conductive, scar tissue can affect the

surrounding, healthy tissue contractions and electrical signaling (Stuart et al. 2016). The

primary goal when using cardiac patches is to provide mechanical support; increasing

wall thickness and limiting infarct expansion. By completing these two goals, the patch

indirectly leads to a third, which is reducing the size of the border region, which leads to

lower incidences of arrhythmias (Stuart et al. 2016, Wan et al. 2013, Amer et al. 2010,

Ong et al. 2013, Olivetti et al. 1990).

Our data suggest that the use of a porcine SIS biomaterial patch does increase

the inflammation response after a 45 minute occlusion-reperfusion myocardial infarction,

either with or without supplemental ibuprofen. Importantly, the addition of the patch to the

infarct area does not appear to result in chronic inflammation, although further studies are

required to conclusively confirm this assessment. Both the PCR and cell counts analyses

showed minimal neutrophil presence and activity in the treatment groups at the end of the

103

7 day period, which is an indication of resolved inflammation. Likewise, the treatments do

not appear to promote excessive or inadequate collagen production, at least within the

inflammation phase. While collagen deposition was noticeable in the patch treatment

groups, both Figures 22 and 23 show that this collagen deposition was limited to the area

immediately surrounding the patch, and therefore is unlikely to increase the risk of

pathophysiology. This is also supported by the majority of the tensile properties of all

groups. The similar elastic moduli and ultimate tensile strength of all infarct groups shows

that the patch’s presence did not significantly affect the tissue turnover in the 7 days post-

infarct, relative to the infarct-only group. As excessive inflammation would have resulted

in high leukocyte infiltration and earlier deposition of collagen, the tensile testing can

provide some comparative information on the state of inflammation (Chen et al. 2013,

Shinde et al. 2014, ). By maintaining the same stiffness and relatively similar protein

composition, the patch allows inflammation to continue while providing the mechanical

support for which it is intended.

It is possible that some of the variations seen between the groups are due to this

mechanical support. Several studies have shown that application of a mechanically

supportive cardiac patch will prevent expansion of the infarct area during wound healing

(Wendel et al. 2014, Zhang et al. 2015, Robinson et al. 2005). A reduction or prevention

of injury expansion in the patch groups are supported by the downregulation of

inflammatory markers at day 7 relative to the infarct only group. It is also worth noting that

IL-RA expression is significantly increased in both treatment groups during the times that

macrophage activity is highest. This suggests that the patch presence may encourage

expression of IL-RA, resulting in slightly faster resolution of inflammation, as IL-RA

104

prevents interleukin action as a receptor antagonist (RA). Faster resolution may be due

to biochemical interactions with the patch itself or, more likely, due to the patch

mechanically limiting the spread of injury and therefore, limiting the inflamed area.

Supporting this conclusion is that both TGF-β and TNF-α are expressed in lower (albeit,

statistically insignificant) levels in both treatment groups relative to infarct only expression.

This suggests that the macrophage behavior is directed towards inflammation resolution,

and not continuation as is often seen with implanted biomaterials (Wendel et al. 2014,

Zhang et al. 2015, Robinson et al. 2005, Ricciotti et al. 2011).

Interestingly, these data also suggest that low-dose NSAID treatment has minimal

effect, if any, on the wound healing process in myocardial infarction. NSAID

administration post-MI has been reviewed in a number of studies, with non-conclusive

results (Leshnower et al. 2006, Vuohelainen et al. 2016, Olivetti et al. 1990, Vargas-

Lorenzo et al. 2013). The data discussed here is supported by the body of work that

suggests that low doses of NSAIDs do not significantly alter inflammation in an injury as

severe as MI, but may assist more so in pain modification and patient comfort (Olivetti et

al. 1990, Vuohelainen et al. 2016). Importantly, these data do not suggest that ibuprofen

results in improper inflammation, either excessive or inadequate, and may be safe to use

post-MI.

Overall, this study provides adequate data to conclude that inflammation is

heightened with the porcine SIS patch treatment. However, the data does not support the

conclusion that the inflammatory process was chronic, or otherwise detrimental to moving

towards the proliferation stage of wound healing, at least compared to the infarct alone.

Therefore, we can conclude that while the patch treatment does increase inflammation,

105

the treatment does not appear to affect inflammation in a manner that poses significant

risk to the long-term wound healing process. This is particularly important when compared

to the mechanical aid that the patch has been shown to provide.

106

CHAPTER 5. CONCLUSION AND FUTURE WORK

Conclusions

These studies demonstrate a broad range of how mechanical interventions can

affect soft tissue repair, both mechanically and biochemically. This work served to provide

a crucial next step in multiple fields of biomedical research, that is, the “test engineering”

phase which identifies potential complications with the design solution inside the system.

First, understanding how suturing affects skin mechanics sheds light on how a

mechanical intervention, sutures, can restore one function of skin (coverage of underlying

tissue), yet have unintended effects on the physiology. By assessing the changes in

elastic modulus, failure strength, and failure strain after suturing parallel or perpendicular

to the tensile axis, we were able to ascertain potential failure modes in sutured soft tissues

that differ from native tissue. This work demonstrates that even the most basic

mechanical interventions can have significant effects on soft tissues at the macroscopic

level.

The next study examined if mechanical interventions can affect soft tissues at a

microscopic level while simultaneously investigating a novel tissue engineering strategy

to replace damaged meniscus tissue. As adipose-derived stromal cells (ASCs) are highly

abundant and are derived from the mesenchyme, they are an ideal source for

differentiation towards a meniscus-like phenotype. By utilizing a combination of

biochemical and mechanical stimulation, we were able to alter the phenotype of the ASCs

and encourage gene expression and protein production of key meniscus proteins, such

as collagen I, collagen II, aggrecan, and versican. By exposing the ASCs to cyclic

mechanical tensile strain, we were successful in inducing fibrosis and chondrogenesis via

107

the TGF-β pathway. This clearly illustrated the potential and necessity of exploring

mechanical intervention when assessing tissue engineering and regenerative medicine

treatments.

Finally, we aim to explore how a mechanical intervention can affect the

inflammation, and healing, of an active, mechanically-dependent soft tissue organ, the

heart. By utilizing concepts on the impact of suturing on dynamic soft tissues as well as

the effects of mechanical forces on the cellular level, we designed an experiment to

assess how the addition of a collagen-based cardiac patch can affect the inflammation,

and ultimately the long-term prognosis, of a rat myocardial infarction model. By comparing

the inflammatory response using gene expression, histology, and inflammatory cell

population and density, we can reasonably gauge how this cardiac patch model will affect

inflammation, and therefore wound healing in a rat infarct model.

The implications of this body of work suggest that mechanical interventions will

play a critical and prominent role in many tissue engineering and regenerative medicine

applications. In particular, the future body of work concerning the use of cardiac patches

to attenuate tissue necrosis and scar tissue bulging in myocardial infarction is of particular

importance. After assessing the initial inflammation status in a patch model, it would be

prudent to evaluate the long term impact of the patch. This analysis would consider not

only mechanical benefits from the patch reinforcement, but how the presence of the patch

affects the proliferation and maturation in the infarct. Furthermore, should the results of

future work suggest that the patch model significantly aids in the promotion of wound

healing in a myocardial infarct model, it would be prudent to optimize the size and material

of the patch. Both of these parameters stand to affect the mechanics of the surrounding

108

tissue, including wall thickness, stiffness, and stress shielding. Altering these parameters

may further improve patient outcomes.

Future Directions

Beyond the tasks of developing an ECM structure and developing an anti-

immunogeneic cell source, there is still much work to be done before a reasonable tissue

developed in vitro can be implemented in vivo. One of the most critical components of

most tissues is vasculature, yet very few studies address vasculature when publishing

work about a novel tissue. In fact, a number of highly publicized studies use decellularized

tissues and stem cells to create “whole” organs to much fanfare (Guyette 2016). However,

these organ structures do not contain a vascular network and would therefore fail quickly

if removed from the cell culture media. This is not to criticize these great works, but to

highlight how the typical approach of ECM + cells is lacking some of the most crucial

elements of a functional tissue.

The lack of vasculature is one of the many reasons why connective tissues and

avascular tissues, such as meniscus, are so attractive to tissue engineers. Connective

tissues tend to be lower cell density, placing a higher priority on excellent ECM structure.

This excellent ECM structure can be obtained through decellularization processes, while

the lower cell population can be provided by stem cells.

109

REFERENCES

1. Altara R, Manca M, Sabra R, Eid A, Booz GW, Zouein FA. (2016) Temporal cardiac

remodeling post-myocarial infarction: dynamics and prognostic implications in

personalized medicine. Heart Fail Rev 21(25): 25-47.

2. Amer M, Bead VR, Bathon J, Blumenthal RS, & Edwards DN. (2010) Use of Nonsteroidal

Anti-Inflammatory Drugs in Patients With Cardiovascular Disease. Cardiology in Review

18:204-212.

3. Arendt EA, Miller LE, & Block JE (2014) Early Knee Osteoarthritis Management Should

First Address Mechanical Joint Overload. Orthop Rev (Pavia) 6: 5188.

4. Baker BM, Nathan AS, Huffman Russell G, & Mauck RL. (2009) Tissue engineering with

meniscus cells derivced from surgical debris. Osteoarthritis and Cartilage 17: 336-345.

5. Baptista LS, Silva KR, Pedrosa CSG, Amaral RJ, Belizario JV, Borojevic R, & Granjeiro

JM. (2013) Bioengineered cartilage in a scaffold-free method by human cartilage-derived

progenitor cells: a comparison with human adipose-derived mesenchymal stromal cells.

Artificial Organs 37(12): 1068-1075.

6. Beck EC, Barragan M, Tadros MH, Gehrke SH, Detamore MS. (2016) Approaching the

compressive modulus of articular cartilage with a decellularized cartilage-based hydrogel.

Acta Biomaterialia. 38: 94-105

7. Boag S, Andreano E, Spyridopoulos I. (2016) Lymphocyte Communication in Myocardial

Ischemia/Reperfusion Injury. Article In Press

8. Bonaventura A, Montecucco F, & Dallegri F. (2016) Cellular recruitment in myocardial

ischaemia/ reperfusion injury. Eur J Clin Invest 46 (6): 590-601.

110

9. Boza P, Ayala P, Vivar R, Humeres C, Cacere FT, Munoz C, Garcia L, Hermoso MA, &

Diaz-Araya G. (2016). Expression and function of toll-like receptor 4 and inflammasomes

in cardiac fibroblasts and myofibroblastsL IL-1β synthesis, secretion, and degradation.

Mol Immunol. 74: 96-105.

10. Bouyarmane H, Beaufils P, Pujol N, Bellemans J, Roberts S, & Spalding T. (2014)

Polyurethane scaffold in lateral meniscus segmental defects: clinical outcomes at 24

months follow-up. Orthop Traumatol Surg Res 100:153-157.

11. Brophy RH, & Matava MJ. (2012) Surgical options for meniscal replacement. J Am Acad

Orthop Surg 20: 265.

12. Buma P, Ramrattan NN, van Tienen TG, & Veth RPH. (2004) Tissue engineering of the

meniscus. Biomaterials 25(1): 1523-1532.

13. Byrne EM, Farrell E, McMahon LA, Haugh MG, O'Brien FJ, Campbell VA. (2008) Gene

expression by marrow stromal cells in a porous collagen-glycosaminoglycan scaffold is

affected by pore size and mechanical stimulation. J Mater Sci Mater Med 19:3455-3463.

14. Caimi G, Lo Presti R, Canino B, Ferrera E, & Hopps E. (2015) Behaviour of the neutrophil

to lymphocyte ratio in young subjects with acute myocardial infarction. Clinical

Hemorheology and Microcirculation. Article In Press.

15. Capek L, Jacquet E, Dzan L, Simunek A. (2012) The analysis of forces needed for the

suturing of elliptical skin wounds. Med Biol Eng Comput. 50(2):193-8.

16. Carrino DA, Calabro A, Darr AB, Dours-Zimmermann MT, Sandy JD, Zimmermann DR,

Sorrell JM, Hascall VC, & Caplan AI. Age-related differences in human skin

proteoglycans. Glycobiology 2: 257-268.

111

17. Case N, Thomas J, Xie Z, Sen B, Styner M, & Rowe D. (2013) Mechanical input restrains

PPARγ2 expression and action to preserve mesenchymal stem cell multipotentiality.

Bone 52: 454-464.

18. Chanda A, Unnikrishnan V, Flynn Z, Lackey K. (2016) Experimental study on tissue

phantoms to understand the effect of injury and suturing on human skin mechanical

properties. Proceedings of the Institution of Mechanical Engineers part H. 231.

19. Chen W, & Frangogiannis NG. (2013) Fibroblasts in post-infarction inflammation and

cardiac repair. Biochimica et Biophysica Acta 1833: 945-953.

20. Cheung HS, Huang CY, & Pelaez D. (2009) Cyclic compression maintains viability and

induces chondrogenesis of human mesenchymal stem cells in fibrin gel scaffolds. Stem

Cells and Development 18(1): 93-100.

21. Chia HN, & Hull ML. (2008) Compressive Moduli of the Human Meniscus in the Axial and

Radial Direction at Equilibrium and at Physiological Strain Rate. Journal of Orthopedic

Research 8: 951-956.

22. Chu B, & Deshmukh, N. (1989) The lack of effect of pentoxifylline on random skin flap

survival. Plastic and Reconstructive Surgery 83: 315-318.

23. Connelly JT, Vanderploeg EJ, Mouw JK, Wilson CG, & Levenston ME. (2010) Tensile

Loading Modulates Bone Marrow Stromal Cell Differentiation and the Development of

Engineered Fibrocartilage Constructs. Tissue Eng A 16: 1913-1923.

24. Corr, DT, & Hart DA. (2011) Biomechanics of Scar Tissue and Uninjured Skin. Advances

in Wound Care (2)2: 37-43.

112

25. D’Amore A, Yoshizumi T, Luketich SK, Wolf MT, Gu X, Cammarata M, Hoff R, Badylak

SF, Wagner WR. (2016) Bi-layered polyurethane- Extracellular matrix cardiac patch

improves ischemic ventricular wall remodeling in a rat model. Biomaterials 107: 1-14.

26. Edwards D, & Elson R. (1995) Skin closure using nylon and polydioxanone: a comparison

of results. Journal of the Royal College of Surgeons of Edinburgh 40: 342-353.

27. Feng L, Moore XL, Dart AM, Wang LM. (2015) Systemic inflammatory response following

acute myocardial infarction. J Geriatr Cardiol. 12:305-312.

28. Frangogiannis NG, Smith CW, Entman ML. (2002) The inflammatory response in

myocardial infarction. Cardiovascular Research 53: 31-47

29. Galie PA, Khalid N, Carnahan KE, Westfall MV, & Stegemann JP. (2013) Substrate

stiffness affects sarcomere and costamere structure and electrophysiological function of

isolated adult cardiomyocytes. Cardiovascular Physiology 22(1): 219-227.

30. Geissman F, Gordon S, Hume DA, Mowat AM, & Randolph GJ. (2010) Unraveling

mononuclear phagocyte heterogeneity. Nature Reviews Immunology 10(6): 453-460.

31. Gerdisch MW, Shea RJ, Barron MD.(2014). Clinical experience with CorMatrix

extracellular matrix in the surgical treatment of mitral valve disease. J Thorac Cardiovasc

Surg. 148:1370-8.

32. Grady E, Elder S, Ryan P, Swiderski C, & Rashmir-Raven A. (2009) Biomechanical and

molecular characteristics of hereditary equine regional dermal asthenia in Quarter

Horses. Veterinary Dermatology 20: 591-599.

33. Greenwald D, Shumway S, Albear P, & Gottlieb L. (1994) Mechanical Comparison of 10

Suture Materials before and after in Vivo Incubation. Journal of Surgical Research 56:

372-377.

113

34. Giroux M, & Descoteaux A. (2000). Cyclooxygenase-2 Expression in Macrophages.

Modulation by Protein Kinase C-α. J Immunol 165: 3985-3991.

35. Godwin JW, Pinto AR, & Rosenthal NA. (2016) Chasing the recipe for a pro-regenerative

immune system. Sem Cell Dev Biol. Article in Press.

36. Harries RL, Bosanquet DC, Harding KG. (2016) Wound bed preparation: TIME for an

update. International Wound Journal. Article In Press.

37. Hasan J, Fisher J, & Ingham, E. (2014) Current strategies in meniscal regeneration. J

Biomed Mater Res B Appl Biomater 102: 619.

38. Hatsushika D, Muneta T, Nakamura T, Horie M, Koga H, & Nakagawa, Y. (2014)

Repetitive allogeneic intraarticular injections of synovial mesenchymal stem cells promote

meniscus regeneration in a porcine massive meniscus defect model. Osteoarthritis

Cartilage 22: 941-950.

39. Higashioka MM, Chen JA, Hu JC, & Athanasiou KA. (2014) Building an anisotropic

meniscus with zonal variations. Tissue Eng Part A 20: 294.

40. Holzapfel, GA. (2000) Biomechanics of Soft Tissue. Computational Biomechanics 7(1):

1-15.

41. Ibrahim A, Ayeni OA, Hughes KB, Lee BT, Slavin SA, Lin SJ. (2013). Acellular Dermal

Matrices in Breast Surgery. Annals of Plastic Surgery 70(6):732-740.

42. Iyer RP, Jung M, Lindsey ML. (2016) MMP-9 signaling in the left ventricle following

myocardial infarction. Am J Physiol Heart Circ Physiol 311: H190-H198.

43. Jansen LA, De Caigny P, Guay N, Lineaweaver WC, Shokrollahi K. (2013) The Evidence

Base for the Acellular Dermal Matrix AlloDerm. Ann Plast Surg 70: 587-594.

114

44. Jugdutt BI, Halliday I, & Uwiera RE. (2007) Therapeutic drugs during healing after

myocardial infarction modify infarct collagens and ventricular distensibility at elevated

pressures. Mol Cell Biochem 304: 79-91.

45. Kapellos TS, Iqbal AJ. (2016) Epigenetic control of macrophage polarisation and soluble

mediator gene expression during inflammation. Mediators of Inflammation. 6(59): 1-15.

46. Kessler D, Dethlefsen S, Haase I, Plomann M, Hirche F, & Krieg, T. (2001) Fibroblasts

in mechanically stressed collagen lattices assume a "synthetic" phenotype. J Biol Chem

276: 36575-36585.

47. Khani MM, Tafazzoli-Shadpour M, Goli-Malekabadi Z, Haghighipour N. Mechanical

characterization of human mesenchymal stem cells subjected to cyclic uniaxial strain and

TGF-B1. J Mech Behav Biomed Mater 43:18-25. 2015.

48. Kirkby NS, Tesfai A, Ahmetaj-Shala B, Gashaw HH, Sampaio W, Etelvino G, Leao NM,

Santos RA, Mitchell JA. Ibuprofen arginate retains eNOS substrate activity and reverses

endothelial dysfunction: implications for the COX2/ADMA axis. The FASEB Journal. 30:1-

7.

49. Kisiday JD, Frisbie DD, McIlwraith CW, & Grodzinsky AJ. (2009) Dynamic compression

stimulates proteoglycan synthesis by mesenchymal stem cells in the absence of

chondrogenic cytokines. Tissue Eng Part A 15: 2817-2824.

50. Kokai LE, Marra KJ, & Rubin JP. (2014) Adipose stem cells: biology and clinical

applications for tissue repair and regeneration. Translational Research 163(4): 399-408.

51. Kong J, Grando SA, & Li YC. (2006) Regulation of IL-1 Family Cytokines IL-1α, IL-1

Receptor Antagonist, and IL-18 by 1,25 Dihydroxyvitamin D3 in Primary Keratinocytes.

The Journal of Immunology 176: 3780-3787

115

52. Kun M, Titan AL, Stafford M, Zheng CH, & Levenston ME. (2012) Variations in

chondrogenesis of human bone marrow-derived mesenchymal stem cells in

fibrin/alginate blended hydrogels. Acta Biomaterialia 8(10): 3754-3764.

53. Lam MT, Huang YC, Birla RK, & Takayama S. (2009) Microfeature guided skeletal

muscle tissue engineering for highly organized 3-dimensional free-standing constructs.

Biomaterials 30(6):1150-1155.

54. Lam MT, Nauta A, Meyer NP, Wu JC, & Longaker MT. (2013) Effective Delivery of Stem

Cells Using an Extracellular Matrix Patch Results in Increased Cell Survival and

Proliferation and Reduced Scarring in Skin Wound Healing. Tissue Engineering A 19:

738-747.

55. Lam MT, Wu JC (2012). Biomaterial applications in cardiovascular tissue repair and

regeneration. Expert Rev Cardiovasc Ther. 10(8): 1039-49.

56. Langton AK, Graham HK, McConnell JC, Sherratt MJ, Griffiths CEM, Watson REB.

(2017) Organisation of the dermal matrix impacts the biomechanical properties of skin.

British Journal of dermatology. Article in Press.

57. Leshnower BG, Sakamoto H, Zeeshan A, Parish LM, & Gorman RC. (2006) Role of

acetaminophen in acute myocardial infarction. Heart and Circulatory Physiology 290:

H2424-2431.

58. Levi K, Ichiryu K, Kefel P, Keller J, Grice J, Belson O, Storne E, Safa B. (2016) Mechanics

of wound closure: emerging tape-based would closure technology vs. traditional methods.

Cureus. 8(10):86.

116

59. Li Z, Kupcsik L, Yao SJ, Alini M, Stoddart M. Mechanical load modulates chondrogenesis

of human mesenchymal stem cells through the TGF-B pathway. J Cell Mol Med. 14:1338-

1346. 2010.

60. Lister Z, Rayner KJ, Suuronen EJ. (2016) How Biomaterials Can Influence Various Cell

Types in the Repair and Regeneration of the Heart after Myocardial Infarction. Front

Bioeng Biotechnol (4):62.

61. Longo UG, Rizzello G, Berton A, Fumo C, Battaglia G, & Khan WS. (2013) A review of

preclinical and clinical studies using synthetic materials for meniscus replacement. Curr

Stem Cell Res Ther 8: 438-443.

62. Makris EA, Hadidi P, & Athanasious A. (2004) The knee meniscus: Structure-function,

pathophysiology, current repair techniques, and prospects for regeneration. Biomaterials

32: 7411-7431.

63. Mandal BB, Park SH, Gil ES, & Kaplan DL. (2011) Stem cell-based meniscus tissue

engineering. Tissue Eng Part A. 17: 2749-2761.

64. McDermott ID, Masouros SD, & Amis AA. (2008) Biomechanics of the meniscus of the

knee. Current Orthopaedics 22: 193-201.

65. Meier EM, Lam MT. (2016) Role of Mechanical Stimulation in Stem Cell Differentiation.

JSM Biotechnology and Biomedical Engineering. Article in Press.

66. Meier EM, Wu B, Siddiqui A, Tepper D, Lam MT. (2016) Mechanical Stimulation

Increases RNA-level expression of Knee Meniscus Genes in Adipose-Derived Stromal

Cells. Plastic and Reconstructive Surgery. 16;4(9):e864.

117

67. Mewhort HEM, Turnball JD, Meijndert HC, Ngu JMC, Fedak PWM. (2014) Epicardial

infarct repair with basic fibroblast growth factor-enhanced CorMatrix-ECM biomaterial

attenuates postischemic cardiac remodeling. J Thorac Cardiovasc Surg 147:1650-9.

68. Mikic B, Johnson TL, Chhabra AB, Schalet BJ, Womng M, & Hunziker EB. (2000)

Differential effects of embryonic immobilization on the development of fibrocartilaginous

skeletal elements. Journal of Rehabilitation Research and Development 37(2): 127-133.

69. Miller, K. (2000) Biomechanics of soft tissues. Medical Science Monitor 6: 158-167.

70. Naal FD, Schauwecker J, Steinhauser E, Milz S, von Knoch F, & Mittelmeier W. (2008)

Biomechanical and immunohistochemical properties of meniscal cartilage after high

hydrostatic pressure treatment. J Biomed Mater Res B Appl Biomater 87: 19-25.

71. Nah DY, Rhee MY. (2009) The Inflammatory Response and Cardiac Repair After

Myocardial Infarction. The Korean Society of Cardiology. 39:393-398.

72. Nerurkar NL, Han W, Mauck RL, & Elliott DM. (2011) Homologous structure-function

relationships between native fibrocartilage and tissue engineered from MSC-seeded

nanofibrous scaffolds. Biomaterials 32: 461-468.

73. Nishimuta JF, & Levenston ME. (2012) Response of cartilage and meniscus tissue

explants to in vitro compressive overload. Osteoarthritis and Cartilage 5(20): 422-429.

74. Oh SJ, Kim Y. (2011) Combined AlloDerm and thin skin grafting for the treatment of

postburn dyspigmented scar contracture of the upper extremity. Journal of Plastic

Reconstructive and Aesthetic Surgery 64, 229-233.

75. Okuno M, Muneta T, Koga H, Ozeki N, Nakagawa Y, & Tsuji K. (2014) Meniscus

regeneration by syngeneic, minor mismatched, and major mismatched transplantation of

synovial mesenchymal stem cells in a rat model. J Orthop Res 32: 928-936.

118

76. Olsen AM, Fosbol EL, Lindhardsen J, Folke F, Charlot M, Selmer C, Lambers M, Olsesen

JB, Kober L, Hansen PR, Torp-Pedersen C, Gislason G. (2011). Duration of Treatment

With Nonsteroidal Anti-Inflammatory Drugs and Impact on Risk of Death and Recurrent

Myocardial Infarction in Patients with Prior Myocardial Infarction. Circulation. 123: 2226-

2235.

77. Olivetti G, Capasso JM, Sonnenblick EH, & Anversa P. (1990) Side-to-Side Slippage of

Myocytes Participates in Ventricular Wall Remodeling Acutely After Myocardial Infarction

in Rats. Circulation Research 6: 23-34.

78. Ong HT, Ong LM, Tan TE, & Chean KY. (2013) Cardiovascular Effects of Common

Analgesics. Med J Malaysia 68: 189- 176.

79. Otsubo H. (2016) Repair of Acute Patellar Tendon Rupture Augmented with Strong

Sutures. J Knee Surg. 1538-8506.

80. Pak J, Lee JH, & Lee SH. (2014) Regenerative repair of damaged meniscus with

autologous adipose tissue-derived stem cells. Biomed Res Int. 3: 436029-436039.

81. Pak J, Lee JH, & Lee SH. (2014) Regenerative repair of damaged meniscus with

autologous adipose tissue-derived stem cells. BioMed Research International.

82. Park JH, Ushida T, & Akimoto T. (2013) Control of cell differentiation by mechanical

stress. J Phys Fitness Sports Med 2(1): 49-62.

83. Patel P, Parikh M, Shah H, Gandhi T. (2016) Inhibition of RhoA/Rho kinase by ibuprofen

exerts cardioprotective effect on isoproterenol induced myocardial infarction in rats.

European Journal of Pharmacology. 791: 91-98.

84. Patrono C. Cardiovascular Effects of Nonsteroidal Anti-Inflammatory Drugs. Curr Cardiol

Rep. 18:25-32.

119

85. Pelaez D, Huang CC, & Cheung HS. (2009) Cyclic compression maintains viability and

induces chondrogenesis of human mesenchymal stem cells in fibrin gel scaffolds. Stem

Cells and Development 1: 93-102.

86. Perea-Gil I, Prat-Vidal C, & Bayes-Genis A. (2015) In vivo experience with natural

scaffolds for myocardial infarction: the times they are a-changin’. Stem Cell Res Ther 6:

248-256

87. Petri M, Ufer K, Toma I, & Becher C. (2012) Effects of perfusion and cyclic compression

on in vitro tissue engineered meniscus implants. Knee Surg Sports Traumatal Arthrosc

20: 223-231.

88. Przyklenk K. (2013). Reduction of Myocardial Infarct Size with Ischemic “Conditioning”:

Physiologic and Technical Considerations. Anesth Analg. 117:891-901.

89. Przyklenk K. (2015). Ischemic conditioning: pitfalls on the path to clinical translation.

British Journal of Pharmacology. 172: 1961-1973.

90. Przyklenk K, Dong Y, Undyala VV, Whittaker P. (2012) Autophagy as a therapeutic target

for ischaemia/reperfusion injury? Concepts, controversies, and challenges.

Cardiovascular Research. 94: 197-205.

91. Qazi SM, Sindby EJ, Norgaard MA. (2015) Ibuprofen-a Safe Analgesic During Cardiac

Surgery Recovery? A Randomized Controlled Trial. J Cardiovasc Thorac Res. 7(4): 141-

148.

92. Ricciotti E, & FitzGerald GA. (2011) Prostaglandins and Inflammation. Arterioscler

Thromb Vasc Biol 31(5): 986-1000.

120

93. Rich L and Whittaker P. (2005) Collagen and Picrosirius Red staining: a polarized light

assessment of fibrillary hue and spatial distribution. Journal of Morphological Sciences

22: 97-104.

94. Robinson KA, Li J, Mathison M, Redkar A, Cui J, Chronos NAF, Matheny RG, & Badylak

SF. (2005) Extracellular Matrix Scaffold for Cardiac Repair. Circulation 112: I-135-I-143.

95. Roeder C, & Driesch P. (1999) Psoriatic erythroderma and bullous pemphigoid treated

successfully with acitretin and azathioprine. European Journal of Dermatology 9: 537-

539.

96. Rongen JJ, van Tienen TG, van Bochove B, Grijpma DW, & Buma P.(2014) Biomaterials

in search of a meniscus substitute. Biomaterials 35: 3527.

97. Rothrauff BB, Numpaisal PO, Lauro BB, Alexander PG, Debski RE, Musahl V, Tuan RS.

(2016) Augmented repair of radial meniscus tear with biomimetic electrospun scaffold: an

in vitro mechanical analysis. J Exp Orthop 3(1):23-30.

98. Roussis PC, Giannakopoulos AE, Charalambous HP. (2015) Suture line response of

end-to-side anastomosis: A stress concentration methodolody. Cardiovasc Eng Techno

6(1):36-48.

99. Saha S, Ji L, de Pablo JJ, Palecek SP. Inhibition of human embryonic stem cell

differentiation by mechanical strain. J Cell Physiol. 206:126-137. 2006.

100. Saha S, Ji L, de Pablo JJ, Palecek SP. TGFB/Activin/Nodal Pathway in Inhibition of

Human Embryonic Stem Cell Differentiation by Mechanical Strain. Biophys J. 94:4123-

4133. 2008.

121

101. Saulis A, Lautenschlager E, & Mustoe T. (2002) Biomechanical and viscoelastic

properties of skin, SMAS, and composite flaps as they pertain to rhytidectomy 110: 177-

183.

102. Schleip R, & Muller DG. (2013) Training principles for fascial connective tissues:

Scientific foundation and suggested practical applications. Journal of Bodywork and

Movement Therapies 17(1): 103-155.

103. Schüttler KF, Pöttgen S, Getgood A, Rominger MB, Fuchs-Winkelmann S, & Roessler

PP. (2014) Improvement in outcomes after implantation of a novel polyurethane meniscal

scaffold for the treatment of medial meniscus deficiency. Knee Surg Sports Traumatol

Arthrosc 7: 1929-1935.

104. Scotti C, Hirschmann MT, Antinolfi P, Martin I, & Peretti GM. (2013) Meniscus repair

and regeneration: review on current methods and research potential. Eur Cell Mater 26:

150.

105. Selvi E, Manganelli S, Catenaccio M, De Stefano R, Frati E, Cucini S, & Marcolongo R.

(2000) Diff Quik staining method for detection and identification of monosodium urate and

calcium phosphate crystals in synovial fluids. Ann Rheum Dis 60:194-198.

106. Sen B, Guilluy C, Xie Z, Case N, Styner M, Thomas J, Oguz I, Rubin C, Burridge K,

Rubin J. Mechanically Induced Focal Adhesion Assembly Amplifies Anti-Adipogenic

Pathways in Mesenchymal Stem Cells. Stem Cells. 29:1829-1836. 2011.

107. Serpooshan V, Zho M, Metzler SA, Wei K, Shah PB, Wang A, Mahmoudi M, Malkovskiy

AV, Rajadas J, Butte MJ, Bernstein D, Ruiz-Lozano P. (2013). The effect of

bioengineered acellular collagen patch on cardiac remodeling and ventricular function

post myocardial infarction. Biomaterials 34:9048-9055.

122

108. Shahabiour F, Banach M, Johnston TP, Pirro M, Sahebkar. (2016) Novel approaches

toward the generation of bioscaffolds as a potential therapy in cardiovascular tissue

engineering. International journal of cardiology. 228 319-326.

109. Shinde AV, & Frangogiannis NG. (2014) Fibroblasts in myocardial infarction: A role in

inflammation and repair. Journal of Molecular and Cellular Cardiology 70:74-82.

110. Shwartz Y, Blitz E, & Zelzer E. (2013) One load to rule the all: Mechanical control of the

musculoskeletal system in development and aging. Differentiation. 86(3): 104-111.

111. Silverman, JF. Inflammatory and neoplastic processes of the lung: differential diagnosis

and pitfalls in FNA biopsies. Diagnostic Cytopathy 13(5): 448-462.

112. Son M, Goodman SB, Chen W, Hargreaves BA, Gold GE, & Levenston ME. (2013)

Regional variation in T1p and T2 times in osteoarthritic human menisci: correlation with

mechanical properties and matrix composition. Osteoarthritis and Cartilage 6(21): 796-

805.

113. Stuart SD, De Jesus NM, Lindsey ML. (2016) Ripplinger, CM. The crossroads of

inflammation, fibrosis, and arrhythmia following myocardial infarction. Journal of

Molecular and Cellular Cardiology 91: 114-122.

114. Stuart SDF, De Jesus NM, Lindsey ML, & Ripplinger CM. (2016) The crossroads of

inflammation, fibrosis, and arrhythmia following myocardial infarction. J Mol Cell Car 91:

114-122.

115. Sutherland AJ, Converse GL, Hopkins RA, Detamore MS. 2015. The bioactivity of

cartilage extracellular matrix in articular cartilage regeneration. Adv Healthc Mater. 4(1):

29-39.

123

116. Tan MY, Zhi W, Wei RQ, Huang YC, Zhou KP, Tan B, Deng L, Luo JC, Li XQ, Xie HQ,

Yang ZM. (2009). Repair of infarcted myocardium using mesenchymal stem cell seeded

small intestinal submucosa in rabbits. Biomaterials. 30:3234-3240.

117. Teramura T, Takehara T, Onodera Y, Nakagawa K, Hamanishi C, Fukuda K.

Mechanical stimulation of cyclic tensile strain induces reduction of pluripotent related

gene expressions via activation of Rho/ROCK and subsequent decreasing of AKT

phosphorylation in human induced pluripotent stem cells. Biochem Biophys Res Commun

417:836-841. 2012

118. Thambyah A, Nather A, & Goh J. (2006) Mechanical properties of articular cartilage

covered by the meniscus. Osteoarthritis and Cartilage 14: 580-588.

119. Toeg HD, Tiwari-Pandey R, Seymour R, Ahmadi A, Crowe S, Vulesevic B, Suuronen

EJ, Ruel M. (2013). Injectable Small Intestine Submucosal Extracellular Matrix in an

Acute Myocardial Infarction Model. Ann Thorac Surg 96:1686-94.

120. Turer AT, & Hill JA. (2010) Pathogenesis of Myocardial Ischemia-Reperfusion Injurt and

Rationale for Therapy. Am J Cardiol 106(3): 360- 368

121. Van den Berg TK, Puklavec MJ, Barclay AN, Dijkstra CD. (2001) Monoclonal antibodies

against rat leukocyte surface antigens. Immunological Reviews. 183:109-116.

122. Van Hout GPJ, van Solinge WW, Gijsberts CM, & Hoefer IE. (2015). Elevated mean

neutrophil volume represents altered neutrophil composition and reflects damage after

myocardial infarction. Basic Res Cardiol 110: 58-69

123. Vangsness CT Jr., Farr J 2nd, Boyd J, Dellaero DT, Mills CR, & LeRoux-Williams M.

(2014) Adult human mesenchymal stem cells delivered via intra-articular injection to the

124

knee following partial medial meniscectomy: a randomized, double-blind, controlled

study. J Bone Joint Surg Am 96: 90-98.

124. Vargas-Lorenzo C, Riera-Guardia N, Calingaert B, Castellsague J, Salvo F, Nicotra F,

Sturkenboom M, & Perez-Gutthann S. (2013) Myocardial infarction and individual

nonsteroid anti-inflammatory drugs meta-analysis of observational studies.

Pharamcoepidemiol Drug Saf 22(6): 559-570.

125. Vesentini S, Redalli A, & Montevecchi F. (2003) Skin Nanostructural Features

Determine Suture Biomechanics. Transactions on Nanobioscience 2: 79-87.

126. Voorhees AP, DeLeon-Pennell KY, Ma Y, Halade GV, Yabluchanskiy A, Iyer RP, Flynn

E, Cates CA, Lindsey ML, Han HC. (2015). Building a better infarct: Modulation of

collagen cross-linking to increase infarct stiffness and reduce left ventricular dilation post-

myocardial infarction. Journal of Molecular and Cellular Cardiology. 85:229-239.

127. Vuohelainen V, Paavonen T, Hamalainen M, Moilanen E, & Mennander AA. (2014) C4d

Deposition Reveals Myocardial Infarction After Cardiac Arrest- Experimental Study. Adv

Clin Exp Med Article in Press.

128. Wan E, Yeap XY, Dehn S, Terry R, Novak M, Zhang S, & Thorp EB. (2013) Enhanced

Efferocytosis of Apoptotic Cardiomyocytes Through Myelod-Epithelial-Reproductive

Tyrosine Kinase Links Acute Inflammation Resolution to Cardiac Repair After Infarction.

Circulation Research 113: 1004-1012.

129. Wan F, Letavernier E, Jourdan Le Saux C, Houssani A, Abid S, Czibik G, Sawaki D,

Marcos E, Dubois-Rande JL, Baud, L, Adnot S, Derumeaux G, & Gellen B. (2015)

Calpastatin Overexpression Impairs Post-infarct Scar Healing in Mice by Compromising

125

Reparative Immune Cell Recruitment and Attachment. Am J Physiol Heart Circ Physiol.

Article In Press.

130. Wan L, Chan Y, Wang Z, Wang W, Schmull S, Dong J, Xue S, Imboden H, Li J. (2017)

Human heart valve- derived scaffold improves cardiac repair in a murine model of

myocardial infarction. Sci Rep. Article in Press.

131. Wan Z, Zheng S, Fan Y, Liu K, Du F, Davey AM, Zhang H, Han W, Xiong C, & Liu W.

(2013) B Cell Activation Is Regulated by the Stiffness Properties of the Substrate

Presenting the Antigens. The Journal of Immunology 190: 4661-4675.

132. Wang A, Kleinerman R, Armstrong A, Fitzmaurice S, Pascucci A, Awasthi S,

Ratnarathorn M, Sivamani R, King T, & Eisen D. (2014) Set-back buried vertical mattress

suturing: results of a randomized blinded trial. Journal of American Academy of

Dermatology 72: 674-680.

133. Wendel JS, Ye L, Zhang P, Tranquillo RT, & Zhang JJ. (2014) Functional

Consequences of a Tissue-Engineered Myocardial Patch for Cardiac Repair in a Rat

Infarct Model. Tissue Engineering A 20: 1325-1335.

134. Wenger MPE, Bozec, L, Horton MA, & Mesquida P. (2007) Mechanical Properties of

Collagen Fibrils. Biophysical Journal 93: 1255-1263.

135. Whittaker P, & Przyklenk K. (2009) Fibrin architecture in clots: A quantitative polarized

light microscopy analysis. Blood Cells Mol Dis 42: 51-56.

136. Xu Y, Malladi P, Chiou M, Bekerman E, Giaccia AJ, Longaker MT. (2007) In vitro

expansion of adipose-derived adult stromal cells in hypoxia enhances early

chondrogenesis. Tissue Eng 13: 2981-2993.

126

137. Yuan X, Arkonac DE, Chao PH, & Vunjak-Novakovic G. (2014) Electrical stimulation

enhances cell migration and integrative repair in the meniscus. Sci Rep 4: 3674.

138. Zhang J (2015) Engineered Tissue Patch for Cardiac Cell Therapy. Curr Treat Options

Cardio Med 17: 37-48.

139. Zhu F, Dong L, Jin X, Jiang B, Kalra A, Shen M, & Yang K. (2015) Testing and modeling

the responses of Hybrid III Crash-Dummy Lower Extremity under high-speed vertical

loading. Stapp Car Crash Journal 59: 521-536.

127

ABSTRACT

MECHANICAL INTERVENTIONS IN SOFT TISSUE REPAIR

by

ELIZABETH MEIER

May 2017

Advisor: Dr. Mai Lam

Major: Biomedical Engineering

Degree: Doctor of Philosophy

This body of work sets to investigate some of these mechanical interventions that

are designed to promote wound healing, repair, or even replace an injured tissue. By

investigating three separate tissues and three separate mechanical interventions, we can

draw conclusions about the implications of including mechanical interventions in

biomedical research and clinical treatments. The use of sutures to close wounds is highly

common, however the effects of sutures on the tensile mechanics of human skin are

largely unknown. To evaluate how sutures may affect uniaxial tensile mechanics, human

skin samples were sutured and loaded in tension in multiple orientations. The data

suggested that the sutured skin had a lower fracture strength and higher elastic modulus

than the intact skin, particularly when loaded in-line with the injury. Next, the inflammatory

effects of a decellularized ECM patch in a myocardial infarction model were analyzed. A

commercially available decellularized material, porcine small intestine submucosa, was

evaluated as a patch treatment in a rat myocardial infarction model, a treatment that is

common in cardiac research. As anticipated, the addition of the patch in the injury area

increased local inflammation as indicated by gene expression and leukocyte population

128

and density. However, the patch did not appear to extend the inflammation response nor

affect the response in a manner that would suggest hindrance to wound healing. Thirdly,

a unique biochemical and mechanical approach was used to direct human adipose stem

cells to differentiate towards a meniscus-like phenotype. By using a variety of media

formulations and a variation of uniaxial tensile parameters, a protocol to maximize

meniscus gene expression was concluded. A chondrogenic media formulation with 10%

uniaxial strain at 1 Hz for 3 hours was found to have the greatest increase in meniscus

gene expression of all of the parameters tested. Together, each of these individual works

contributes to the conclusion that mechanical interventions can have a significant impact

on the restructuring, repair, and replacement of soft tissues.

129

AUTOBIOGRAPHICAL STATEMENT

. Elizabeth majored in biomedical engineering at Michigan Technological University

in Houghton, MI. Her work in the healthcare industry and as a professional engineer

fueled her desire to continue her education to better contribute to the current struggles of

the healthcare system. In her spare time, she enjoys training and competing in road races

and triathlons.

PUBLICATIONS

Elizabeth M. Meier MSE, Bin Wu, Aamir Siddiqui MD, Donna Tepper MD, Mai T Lam

PhD. Mechanical Stimulation Increases RNA-level expression of Knee Meniscus Genes

in Adipose-Derived Stromal Cells. Plastic and Reconstructive Surgery. 2016 Sep

16;4(9):e864.

Elizabeth M. Meier MSE, Mai T Lam PhD. Role of Mechanical Stimulation in Stem Cell

Differentiation. JSM Biotechnology and Biomedical Engineering. 2016. Article in Press.

Cameron B Pinnock, MSE; Elizabeth M Meier, MSE; Neeraj N Joshi, MD; Bin Wu, MD;

Customizable Engineered Blood Vessels Using 3D Printed Inserts. Methods. 2016 Apr

15;99:20-7.

Elizabeth M. Meier MSE, Bin Wu, Mai T Lam PhD. The Inflammatory and Mechanical

Effects of Porcine Small Intestine Submucosa Patches on Cardiac Tissue Post-

Myocardial Infarction. Tissue Engineering Part A. Submitted.

Elizabeth M. Meier MSE, Bin Wu, Mai T Lam PhD. Effects of Strain Rate and Suturing on

Mechanical Behavior of Skin. Biomechanics. Submitted